[www-releases] r225843 - Add documentation for 3.5.1

Tom Stellard thomas.stellard at amd.com
Tue Jan 13 14:55:45 PST 2015


Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-cov.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-cov.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-cov.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-cov.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,239 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-cov - emit coverage information — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-profdata - Profile data tool" href="llvm-profdata.html" />
+    <link rel="prev" title="llvm-diff - LLVM structural ‘diff’" href="llvm-diff.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">
+      <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="llvm-profdata.html" title="llvm-profdata - Profile data tool"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-diff.html" title="llvm-diff - LLVM structural ‘diff’"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-cov-emit-coverage-information">
+<h1>llvm-cov - emit coverage information<a class="headerlink" href="#llvm-cov-emit-coverage-information" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-cov</strong> [options] SOURCEFILE</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong class="program">llvm-cov</strong> tool reads code coverage data files and displays the
+coverage information for a specified source file. It is compatible with the
+<tt class="docutils literal"><span class="pre">gcov</span></tt> tool from version 4.2 of <tt class="docutils literal"><span class="pre">GCC</span></tt> and may also be compatible with
+some later versions of <tt class="docutils literal"><span class="pre">gcov</span></tt>.</p>
+<p>To use llvm-cov, you must first build an instrumented version of your
+application that collects coverage data as it runs. Compile with the
+<tt class="docutils literal"><span class="pre">-fprofile-arcs</span></tt> and <tt class="docutils literal"><span class="pre">-ftest-coverage</span></tt> options to add the
+instrumentation. (Alternatively, you can use the <tt class="docutils literal"><span class="pre">--coverage</span></tt> option, which
+includes both of those other options.) You should compile with debugging
+information (<tt class="docutils literal"><span class="pre">-g</span></tt>) and without optimization (<tt class="docutils literal"><span class="pre">-O0</span></tt>); otherwise, the
+coverage data cannot be accurately mapped back to the source code.</p>
+<p>At the time you compile the instrumented code, a <tt class="docutils literal"><span class="pre">.gcno</span></tt> data file will be
+generated for each object file. These <tt class="docutils literal"><span class="pre">.gcno</span></tt> files contain half of the
+coverage data. The other half of the data comes from <tt class="docutils literal"><span class="pre">.gcda</span></tt> files that are
+generated when you run the instrumented program, with a separate <tt class="docutils literal"><span class="pre">.gcda</span></tt>
+file for each object file. Each time you run the program, the execution counts
+are summed into any existing <tt class="docutils literal"><span class="pre">.gcda</span></tt> files, so be sure to remove any old
+files if you do not want their contents to be included.</p>
+<p>By default, the <tt class="docutils literal"><span class="pre">.gcda</span></tt> files are written into the same directory as the
+object files, but you can override that by setting the <tt class="docutils literal"><span class="pre">GCOV_PREFIX</span></tt> and
+<tt class="docutils literal"><span class="pre">GCOV_PREFIX_STRIP</span></tt> environment variables. The <tt class="docutils literal"><span class="pre">GCOV_PREFIX_STRIP</span></tt>
+variable specifies a number of directory components to be removed from the
+start of the absolute path to the object file directory. After stripping those
+directories, the prefix from the <tt class="docutils literal"><span class="pre">GCOV_PREFIX</span></tt> variable is added. These
+environment variables allow you to run the instrumented program on a machine
+where the original object file directories are not accessible, but you will
+then need to copy the <tt class="docutils literal"><span class="pre">.gcda</span></tt> files back to the object file directories
+where llvm-cov expects to find them.</p>
+<p>Once you have generated the coverage data files, run llvm-cov for each main
+source file where you want to examine the coverage results. This should be run
+from the same directory where you previously ran the compiler. The results for
+the specified source file are written to a file named by appending a <tt class="docutils literal"><span class="pre">.gcov</span></tt>
+suffix. A separate output file is also created for each file included by the
+main source file, also with a <tt class="docutils literal"><span class="pre">.gcov</span></tt> suffix added.</p>
+<p>The basic content of an llvm-cov output file is a copy of the source file with
+an execution count and line number prepended to every line. The execution
+count is shown as <tt class="docutils literal"><span class="pre">-</span></tt> if a line does not contain any executable code. If
+a line contains code but that code was never executed, the count is displayed
+as <tt class="docutils literal"><span class="pre">#####</span></tt>.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<dl class="option">
+<dt id="cmdoption-a">
+<tt class="descname">-a</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">--all-blocks</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-a" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display all basic blocks. If there are multiple blocks for a single line of
+source code, this option causes llvm-cov to show the count for each block
+instead of just one count for the entire line.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-b">
+<tt class="descname">-b</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">--branch-probabilities</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-b" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display conditional branch probabilities and a summary of branch information.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-c">
+<tt class="descname">-c</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">--branch-counts</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-c" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display branch counts instead of probabilities (requires -b).</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-f">
+<tt class="descname">-f</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">--function-summaries</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-f" title="Permalink to this definition">¶</a></dt>
+<dd><p>Show a summary of coverage for each function instead of just one summary for
+an entire source file.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption--help">
+<tt class="descname">--help</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption--help" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display available options (–help-hidden for more).</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-l">
+<tt class="descname">-l</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">--long-file-names</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-l" title="Permalink to this definition">¶</a></dt>
+<dd><p>For coverage output of files included from the main source file, add the
+main file name followed by <tt class="docutils literal"><span class="pre">##</span></tt> as a prefix to the output file names. This
+can be combined with the –preserve-paths option to use complete paths for
+both the main file and the included file.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-n">
+<tt class="descname">-n</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">--no-output</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-n" title="Permalink to this definition">¶</a></dt>
+<dd><p>Do not output any <tt class="docutils literal"><span class="pre">.gcov</span></tt> files. Summary information is still
+displayed.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-o">
+<tt class="descname">-o</tt><tt class="descclassname">=<DIR|FILE></tt><tt class="descclassname">, </tt><tt class="descname">--object-directory</tt><tt class="descclassname">=<DIR></tt><tt class="descclassname">, </tt><tt class="descname">--object-file</tt><tt class="descclassname">=<FILE></tt><a class="headerlink" href="#cmdoption-o" title="Permalink to this definition">¶</a></dt>
+<dd><p>Find objects in DIR or based on FILE’s path. If you specify a particular
+object file, the coverage data files are expected to have the same base name
+with <tt class="docutils literal"><span class="pre">.gcno</span></tt> and <tt class="docutils literal"><span class="pre">.gcda</span></tt> extensions. If you specify a directory, the
+files are expected in that directory with the same base name as the source
+file.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-p">
+<tt class="descname">-p</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">--preserve-paths</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-p" title="Permalink to this definition">¶</a></dt>
+<dd><p>Preserve path components when naming the coverage output files. In addition
+to the source file name, include the directories from the path to that
+file. The directories are separate by <tt class="docutils literal"><span class="pre">#</span></tt> characters, with <tt class="docutils literal"><span class="pre">.</span></tt> directories
+removed and <tt class="docutils literal"><span class="pre">..</span></tt> directories replaced by <tt class="docutils literal"><span class="pre">^</span></tt> characters. When used with
+the –long-file-names option, this applies to both the main file name and the
+included file name.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-u">
+<tt class="descname">-u</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">--unconditional-branches</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-u" title="Permalink to this definition">¶</a></dt>
+<dd><p>Include unconditional branches in the output for the –branch-probabilities
+option.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-version">
+<tt class="descname">-version</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-version" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display the version of llvm-cov.</p>
+</dd></dl>
+
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-cov</strong> returns 1 if it cannot read input files.  Otherwise, it
+exits with zero.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-profdata.html" title="llvm-profdata - Profile data tool"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-diff.html" title="llvm-diff - LLVM structural ‘diff’"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-diff.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-diff.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-diff.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-diff.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,139 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-diff - LLVM structural ‘diff’ — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-cov - emit coverage information" href="llvm-cov.html" />
+    <link rel="prev" title="llvm-config - Print LLVM compilation options" href="llvm-config.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">
+      <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="llvm-cov.html" title="llvm-cov - emit coverage information"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-config.html" title="llvm-config - Print LLVM compilation options"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-diff-llvm-structural-diff">
+<h1>llvm-diff - LLVM structural ‘diff’<a class="headerlink" href="#llvm-diff-llvm-structural-diff" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong>llvm-diff</strong> [<em>options</em>] <em>module 1</em> <em>module 2</em> [<em>global name ...</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p><strong>llvm-diff</strong> compares the structure of two LLVM modules, primarily
+focusing on differences in function definitions.  Insignificant
+differences, such as changes in the ordering of globals or in the
+names of local values, are ignored.</p>
+<p>An input module will be interpreted as an assembly file if its name
+ends in ‘.ll’;  otherwise it will be read in as a bitcode file.</p>
+<p>If a list of global names is given, just the values with those names
+are compared; otherwise, all global values are compared, and
+diagnostics are produced for globals which only appear in one module
+or the other.</p>
+<p><strong>llvm-diff</strong> compares two functions by comparing their basic blocks,
+beginning with the entry blocks.  If the terminators seem to match,
+then the corresponding successors are compared; otherwise they are
+ignored.  This algorithm is very sensitive to changes in control flow,
+which tend to stop any downstream changes from being detected.</p>
+<p><strong>llvm-diff</strong> is intended as a debugging tool for writers of LLVM
+passes and frontends.  It does not have a stable output format.</p>
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p>If <strong>llvm-diff</strong> finds no differences between the modules, it will exit
+with 0 and produce no output.  Otherwise it will exit with a non-zero
+value.</p>
+</div>
+<div class="section" id="bugs">
+<h2>BUGS<a class="headerlink" href="#bugs" title="Permalink to this headline">¶</a></h2>
+<p>Many important differences, like changes in linkage or function
+attributes, are not diagnosed.</p>
+<p>Changes in memory behavior (for example, coalescing loads) can cause
+massive detected differences in blocks.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-cov.html" title="llvm-cov - emit coverage information"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-config.html" title="llvm-config - Print LLVM compilation options"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-dis.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-dis.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-dis.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-dis.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,143 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-dis - LLVM disassembler — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="opt - LLVM optimizer" href="opt.html" />
+    <link rel="prev" title="llvm-as - LLVM assembler" href="llvm-as.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">
+      <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="opt.html" title="opt - LLVM optimizer"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-as.html" title="llvm-as - LLVM assembler"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-dis-llvm-disassembler">
+<h1>llvm-dis - LLVM disassembler<a class="headerlink" href="#llvm-dis-llvm-disassembler" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong>llvm-dis</strong> [<em>options</em>] [<em>filename</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong>llvm-dis</strong> command is the LLVM disassembler.  It takes an LLVM
+bitcode file and converts it into human-readable LLVM assembly language.</p>
+<p>If filename is omitted or specified as <tt class="docutils literal"><span class="pre">-</span></tt>, <strong>llvm-dis</strong> reads its
+input from standard input.</p>
+<p>If the input is being read from standard input, then <strong>llvm-dis</strong>
+will send its output to standard output by default.  Otherwise, the
+output will be written to a file named after the input file, with
+a <tt class="docutils literal"><span class="pre">.ll</span></tt> suffix added (any existing <tt class="docutils literal"><span class="pre">.bc</span></tt> suffix will first be
+removed).  You can override the choice of output file using the
+<strong>-o</strong> option.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<p><strong>-f</strong></p>
+<blockquote>
+<div>Enable binary output on terminals.  Normally, <strong>llvm-dis</strong> will refuse to
+write raw bitcode output if the output stream is a terminal. With this option,
+<strong>llvm-dis</strong> will write raw bitcode regardless of the output device.</div></blockquote>
+<p><strong>-help</strong></p>
+<blockquote>
+<div>Print a summary of command line options.</div></blockquote>
+<p><strong>-o</strong> <em>filename</em></p>
+<blockquote>
+<div>Specify the output file name.  If <em>filename</em> is -, then the output is sent
+to standard output.</div></blockquote>
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p>If <strong>llvm-dis</strong> succeeds, it will exit with 0.  Otherwise, if an error
+occurs, it will exit with a non-zero value.</p>
+</div>
+<div class="section" id="see-also">
+<h2>SEE ALSO<a class="headerlink" href="#see-also" title="Permalink to this headline">¶</a></h2>
+<p>llvm-as|llvm-as</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="opt.html" title="opt - LLVM optimizer"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-as.html" title="llvm-as - LLVM assembler"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-dwarfdump.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-dwarfdump.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-dwarfdump.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-dwarfdump.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,129 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-dwarfdump - print contents of DWARF sections — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="bugpoint - automatic test case reduction tool" href="bugpoint.html" />
+    <link rel="prev" title="llvm-symbolizer - convert addresses into source code locations" href="llvm-symbolizer.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">
+      <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="bugpoint.html" title="bugpoint - automatic test case reduction tool"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-symbolizer.html" title="llvm-symbolizer - convert addresses into source code locations"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-dwarfdump-print-contents-of-dwarf-sections">
+<h1>llvm-dwarfdump - print contents of DWARF sections<a class="headerlink" href="#llvm-dwarfdump-print-contents-of-dwarf-sections" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-dwarfdump</strong> [<em>options</em>] [<em>filenames...</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-dwarfdump</strong> parses DWARF sections in the object files
+and prints their contents in human-readable form.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<dl class="option">
+<dt id="cmdoption-debug-dump">
+<tt class="descname">-debug-dump</tt><tt class="descclassname">=section</tt><a class="headerlink" href="#cmdoption-debug-dump" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the DWARF section to dump.
+For example, use <tt class="docutils literal"><span class="pre">abbrev</span></tt> to dump the contents of <tt class="docutils literal"><span class="pre">.debug_abbrev</span></tt> section,
+<tt class="docutils literal"><span class="pre">loc.dwo</span></tt> to dump the contents of <tt class="docutils literal"><span class="pre">.debug_loc.dwo</span></tt> etc.
+See <tt class="docutils literal"><span class="pre">llvm-dwarfdump</span> <span class="pre">--help</span></tt> for the complete list of supported sections.
+Use <tt class="docutils literal"><span class="pre">all</span></tt> to dump all DWARF sections. It is the default.</p>
+</dd></dl>
+
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-dwarfdump</strong> returns 0. Other exit codes imply internal
+program error.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="bugpoint.html" title="bugpoint - automatic test case reduction tool"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-symbolizer.html" title="llvm-symbolizer - convert addresses into source code locations"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-extract.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-extract.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-extract.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-extract.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,164 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-extract - extract a function from an LLVM module — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-bcanalyzer - LLVM bitcode analyzer" href="llvm-bcanalyzer.html" />
+    <link rel="prev" title="bugpoint - automatic test case reduction tool" href="bugpoint.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">
+      <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="llvm-bcanalyzer.html" title="llvm-bcanalyzer - LLVM bitcode analyzer"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="bugpoint.html" title="bugpoint - automatic test case reduction tool"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-extract-extract-a-function-from-an-llvm-module">
+<h1>llvm-extract - extract a function from an LLVM module<a class="headerlink" href="#llvm-extract-extract-a-function-from-an-llvm-module" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-extract</strong> [<em>options</em>] <strong>–func</strong> <em>function-name</em> [<em>filename</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong class="program">llvm-extract</strong> command takes the name of a function and extracts
+it from the specified LLVM bitcode file.  It is primarily used as a debugging
+tool to reduce test cases from larger programs that are triggering a bug.</p>
+<p>In addition to extracting the bitcode of the specified function,
+<strong class="program">llvm-extract</strong> will also remove unreachable global variables,
+prototypes, and unused types.</p>
+<p>The <strong class="program">llvm-extract</strong> command reads its input from standard input if
+filename is omitted or if filename is <tt class="docutils literal"><span class="pre">-</span></tt>.  The output is always written to
+standard output, unless the <strong>-o</strong> option is specified (see below).</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<p><strong>-f</strong></p>
+<blockquote>
+<div>Enable binary output on terminals.  Normally, <strong class="program">llvm-extract</strong> will
+refuse to write raw bitcode output if the output stream is a terminal.  With
+this option, <strong class="program">llvm-extract</strong> will write raw bitcode regardless of the
+output device.</div></blockquote>
+<p><strong>–func</strong> <em>function-name</em></p>
+<blockquote>
+<div>Extract the function named <em>function-name</em> from the LLVM bitcode.  May be
+specified multiple times to extract multiple functions at once.</div></blockquote>
+<p><strong>–rfunc</strong> <em>function-regular-expr</em></p>
+<blockquote>
+<div>Extract the function(s) matching <em>function-regular-expr</em> from the LLVM bitcode.
+All functions matching the regular expression will be extracted.  May be
+specified multiple times.</div></blockquote>
+<p><strong>–glob</strong> <em>global-name</em></p>
+<blockquote>
+<div>Extract the global variable named <em>global-name</em> from the LLVM bitcode.  May be
+specified multiple times to extract multiple global variables at once.</div></blockquote>
+<p><strong>–rglob</strong> <em>glob-regular-expr</em></p>
+<blockquote>
+<div>Extract the global variable(s) matching <em>global-regular-expr</em> from the LLVM
+bitcode.  All global variables matching the regular expression will be
+extracted.  May be specified multiple times.</div></blockquote>
+<p><strong>-help</strong></p>
+<blockquote>
+<div>Print a summary of command line options.</div></blockquote>
+<p><strong>-o</strong> <em>filename</em></p>
+<blockquote>
+<div>Specify the output filename.  If filename is “-” (the default), then
+<strong class="program">llvm-extract</strong> sends its output to standard output.</div></blockquote>
+<p><strong>-S</strong></p>
+<blockquote>
+<div>Write output in LLVM intermediate language (instead of bitcode).</div></blockquote>
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p>If <strong class="program">llvm-extract</strong> succeeds, it will exit with 0.  Otherwise, if an error
+occurs, it will exit with a non-zero value.</p>
+</div>
+<div class="section" id="see-also">
+<h2>SEE ALSO<a class="headerlink" href="#see-also" title="Permalink to this headline">¶</a></h2>
+<p>bugpoint</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-bcanalyzer.html" title="llvm-bcanalyzer - LLVM bitcode analyzer"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="bugpoint.html" title="bugpoint - automatic test case reduction tool"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-link.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-link.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-link.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-link.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,163 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-link - LLVM bitcode linker — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-ar - LLVM archiver" href="llvm-ar.html" />
+    <link rel="prev" title="lli - directly execute programs from LLVM bitcode" href="lli.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">
+      <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="llvm-ar.html" title="llvm-ar - LLVM archiver"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="lli.html" title="lli - directly execute programs from LLVM bitcode"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-link-llvm-bitcode-linker">
+<h1>llvm-link - LLVM bitcode linker<a class="headerlink" href="#llvm-link-llvm-bitcode-linker" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-link</strong> [<em>options</em>] <em>filename ...</em></p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-link</strong> takes several LLVM bitcode files and links them together
+into a single LLVM bitcode file.  It writes the output file to standard output,
+unless the <a class="reference internal" href="opt.html#cmdoption-o"><em class="xref std std-option">-o</em></a> option is used to specify a filename.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<dl class="option">
+<dt id="cmdoption-f">
+<tt class="descname">-f</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-f" title="Permalink to this definition">¶</a></dt>
+<dd><p>Enable binary output on terminals.  Normally, <strong class="program">llvm-link</strong> will refuse
+to write raw bitcode output if the output stream is a terminal. With this
+option, <strong class="program">llvm-link</strong> will write raw bitcode regardless of the output
+device.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-o">
+<tt class="descname">-o</tt><tt class="descclassname"> filename</tt><a class="headerlink" href="#cmdoption-o" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the output file name.  If <tt class="docutils literal"><span class="pre">filename</span></tt> is “<tt class="docutils literal"><span class="pre">-</span></tt>”, then
+<strong class="program">llvm-link</strong> will write its output to standard output.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-S">
+<tt class="descname">-S</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-S" title="Permalink to this definition">¶</a></dt>
+<dd><p>Write output in LLVM intermediate language (instead of bitcode).</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-d">
+<tt class="descname">-d</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-d" title="Permalink to this definition">¶</a></dt>
+<dd><p>If specified, <strong class="program">llvm-link</strong> prints a human-readable version of the
+output bitcode file to standard error.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-help">
+<tt class="descname">-help</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-help" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print a summary of command line options.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-v">
+<tt class="descname">-v</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-v" title="Permalink to this definition">¶</a></dt>
+<dd><p>Verbose mode.  Print information about what <strong class="program">llvm-link</strong> is doing.
+This typically includes a message for each bitcode file linked in and for each
+library found.</p>
+</dd></dl>
+
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p>If <strong class="program">llvm-link</strong> succeeds, it will exit with 0.  Otherwise, if an error
+occurs, it will exit with a non-zero value.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-ar.html" title="llvm-ar - LLVM archiver"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="lli.html" title="lli - directly execute programs from LLVM bitcode"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-nm.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-nm.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-nm.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-nm.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,259 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-nm - list LLVM bitcode and object file’s symbol table — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-config - Print LLVM compilation options" href="llvm-config.html" />
+    <link rel="prev" title="llvm-ar - LLVM archiver" href="llvm-ar.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">
+      <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="llvm-config.html" title="llvm-config - Print LLVM compilation options"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-ar.html" title="llvm-ar - LLVM archiver"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-nm-list-llvm-bitcode-and-object-file-s-symbol-table">
+<h1>llvm-nm - list LLVM bitcode and object file’s symbol table<a class="headerlink" href="#llvm-nm-list-llvm-bitcode-and-object-file-s-symbol-table" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-nm</strong> [<em>options</em>] [<em>filenames...</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong class="program">llvm-nm</strong> utility lists the names of symbols from the LLVM bitcode
+files, object files, or <strong class="program">ar</strong> archives containing them, named on the
+command line.  Each symbol is listed along with some simple information about
+its provenance.  If no file name is specified, or <em>-</em> is used as a file name,
+<strong class="program">llvm-nm</strong> will process a file on its standard input stream.</p>
+<p><strong class="program">llvm-nm</strong>‘s default output format is the traditional BSD <strong class="program">nm</strong>
+output format.  Each such output record consists of an (optional) 8-digit
+hexadecimal address, followed by a type code character, followed by a name, for
+each symbol.  One record is printed per line; fields are separated by spaces.
+When the address is omitted, it is replaced by 8 spaces.</p>
+<p>Type code characters currently supported, and their meanings, are as follows:</p>
+<p>U</p>
+<blockquote>
+<div>Named object is referenced but undefined in this bitcode file</div></blockquote>
+<p>C</p>
+<blockquote>
+<div>Common (multiple definitions link together into one def)</div></blockquote>
+<p>W</p>
+<blockquote>
+<div>Weak reference (multiple definitions link together into zero or one definitions)</div></blockquote>
+<p>t</p>
+<blockquote>
+<div>Local function (text) object</div></blockquote>
+<p>T</p>
+<blockquote>
+<div>Global function (text) object</div></blockquote>
+<p>d</p>
+<blockquote>
+<div>Local data object</div></blockquote>
+<p>D</p>
+<blockquote>
+<div>Global data object</div></blockquote>
+<p>?</p>
+<blockquote>
+<div>Something unrecognizable</div></blockquote>
+<p>Because LLVM bitcode files typically contain objects that are not considered to
+have addresses until they are linked into an executable image or dynamically
+compiled “just-in-time”, <strong class="program">llvm-nm</strong> does not print an address for any
+symbol in an LLVM bitcode file, even symbols which are defined in the bitcode
+file.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<dl class="option">
+<dt id="cmdoption-llvm-nm-B">
+<tt class="descname">-B</tt><tt class="descclassname">    (default)</tt><a class="headerlink" href="#cmdoption-llvm-nm-B" title="Permalink to this definition">¶</a></dt>
+<dd><p>Use BSD output format.  Alias for <em class="xref std std-option">--format=bsd</em>.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm-P">
+<tt class="descname">-P</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm-P" title="Permalink to this definition">¶</a></dt>
+<dd><p>Use POSIX.2 output format.  Alias for <em class="xref std std-option">--format=posix</em>.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--debug-syms">
+<tt class="descname">--debug-syms</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-a</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--debug-syms" title="Permalink to this definition">¶</a></dt>
+<dd><p>Show all symbols, even debugger only.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--defined-only">
+<tt class="descname">--defined-only</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--defined-only" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print only symbols defined in this file (as opposed to
+symbols which may be referenced by objects in this file, but not
+defined in this file.)</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--dynamic">
+<tt class="descname">--dynamic</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-D</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--dynamic" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display dynamic symbols instead of normal symbols.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--extern-only">
+<tt class="descname">--extern-only</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-g</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--extern-only" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print only symbols whose definitions are external; that is, accessible
+from other files.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--format">
+<tt class="descname">--format</tt><tt class="descclassname">=format</tt><tt class="descclassname">, </tt><tt class="descname">-f</tt><tt class="descclassname"> format</tt><a class="headerlink" href="#cmdoption-llvm-nm--format" title="Permalink to this definition">¶</a></dt>
+<dd><p>Select an output format; <em>format</em> may be <em>sysv</em>, <em>posix</em>, or <em>bsd</em>.  The default
+is <em>bsd</em>.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm-help">
+<tt class="descname">-help</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm-help" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print a summary of command-line options and their meanings.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--no-sort">
+<tt class="descname">--no-sort</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-p</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--no-sort" title="Permalink to this definition">¶</a></dt>
+<dd><p>Shows symbols in order encountered.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--numeric-sort">
+<tt class="descname">--numeric-sort</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-n</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-v</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--numeric-sort" title="Permalink to this definition">¶</a></dt>
+<dd><p>Sort symbols by address.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--print-file-name">
+<tt class="descname">--print-file-name</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-A</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-o</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--print-file-name" title="Permalink to this definition">¶</a></dt>
+<dd><p>Precede each symbol with the file it came from.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--print-size">
+<tt class="descname">--print-size</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-S</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--print-size" title="Permalink to this definition">¶</a></dt>
+<dd><p>Show symbol size instead of address.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--size-sort">
+<tt class="descname">--size-sort</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--size-sort" title="Permalink to this definition">¶</a></dt>
+<dd><p>Sort symbols by size.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-nm--undefined-only">
+<tt class="descname">--undefined-only</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-u</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-nm--undefined-only" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print only symbols referenced but not defined in this file.</p>
+</dd></dl>
+
+</div>
+<div class="section" id="bugs">
+<h2>BUGS<a class="headerlink" href="#bugs" title="Permalink to this headline">¶</a></h2>
+<blockquote>
+<div><ul class="simple">
+<li><strong class="program">llvm-nm</strong> cannot demangle C++ mangled names, like GNU <strong class="program">nm</strong>
+can.</li>
+<li><strong class="program">llvm-nm</strong> does not support the full set of arguments that GNU
+<strong class="program">nm</strong> does.</li>
+</ul>
+</div></blockquote>
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-nm</strong> exits with an exit code of zero.</p>
+</div>
+<div class="section" id="see-also">
+<h2>SEE ALSO<a class="headerlink" href="#see-also" title="Permalink to this headline">¶</a></h2>
+<p>llvm-dis, ar(1), nm(1)</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-config.html" title="llvm-config - Print LLVM compilation options"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-ar.html" title="llvm-ar - LLVM archiver"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-prof.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-prof.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-prof.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-prof.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,141 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-prof - print execution profile of LLVM program — LLVM 3.4 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.4',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.4 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-config - Print LLVM compilation options" href="llvm-config.html" />
+    <link rel="prev" title="llvm-nm - list LLVM bitcode and object file’s symbol table" href="llvm-nm.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">
+      <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="llvm-config.html" title="llvm-config - Print LLVM compilation options"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-nm.html" title="llvm-nm - list LLVM bitcode and object file’s symbol table"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-prof-print-execution-profile-of-llvm-program">
+<h1>llvm-prof - print execution profile of LLVM program<a class="headerlink" href="#llvm-prof-print-execution-profile-of-llvm-program" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong>llvm-prof</strong> [<em>options</em>] [<em>bitcode file</em>] [<em>llvmprof.out</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong>llvm-prof</strong> tool reads in an <em>llvmprof.out</em> file (which can
+optionally use a specific file with the third program argument), a bitcode file
+for the program, and produces a human readable report, suitable for determining
+where the program hotspots are.</p>
+<p>This program is often used in conjunction with the <em>utils/profile.pl</em>
+script.  This script automatically instruments a program, runs it with the JIT,
+then runs <strong>llvm-prof</strong> to format a report.  To get more information about
+<em>utils/profile.pl</em>, execute it with the <strong>-help</strong> option.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<p><strong>–annotated-llvm</strong> or <strong>-A</strong></p>
+<blockquote>
+<div>In addition to the normal report printed, print out the code for the
+program, annotated with execution frequency information. This can be
+particularly useful when trying to visualize how frequently basic blocks
+are executed.  This is most useful with basic block profiling
+information or better.</div></blockquote>
+<p><strong>–print-all-code</strong></p>
+<blockquote>
+<div>Using this option enables the <strong>–annotated-llvm</strong> option, but it
+prints the entire module, instead of just the most commonly executed
+functions.</div></blockquote>
+<p><strong>–time-passes</strong></p>
+<blockquote>
+<div>Record the amount of time needed for each pass and print it to standard
+error.</div></blockquote>
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p><strong>llvm-prof</strong> returns 1 if it cannot load the bitcode file or the profile
+information. Otherwise, it exits with zero.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-config.html" title="llvm-config - Print LLVM compilation options"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-nm.html" title="llvm-nm - list LLVM bitcode and object file’s symbol table"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2013, LLVM Project.
+      Last updated on 2014-06-16.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-profdata.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-profdata.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-profdata.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-profdata.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,201 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-profdata - Profile data tool — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-stress - generate random .ll files" href="llvm-stress.html" />
+    <link rel="prev" title="llvm-cov - emit coverage information" href="llvm-cov.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">
+      <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="llvm-stress.html" title="llvm-stress - generate random .ll files"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-cov.html" title="llvm-cov - emit coverage information"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-profdata-profile-data-tool">
+<h1>llvm-profdata - Profile data tool<a class="headerlink" href="#llvm-profdata-profile-data-tool" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-profdata</strong> <em>command</em> [<em>args...</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong class="program">llvm-profdata</strong> tool is a small utility for working with profile
+data files.</p>
+</div>
+<div class="section" id="commands">
+<h2>COMMANDS<a class="headerlink" href="#commands" title="Permalink to this headline">¶</a></h2>
+<ul class="simple">
+<li><a class="reference internal" href="#profdata-merge">merge</a></li>
+<li><a class="reference internal" href="#profdata-show">show</a></li>
+</ul>
+</div>
+<div class="section" id="merge">
+<span id="profdata-merge"></span><h2>MERGE<a class="headerlink" href="#merge" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="id1">
+<h3>SYNOPSIS<a class="headerlink" href="#id1" title="Permalink to this headline">¶</a></h3>
+<p><strong class="program">llvm-profdata merge</strong> [<em>options</em>] [<em>filenames...</em>]</p>
+</div>
+<div class="section" id="id2">
+<h3>DESCRIPTION<a class="headerlink" href="#id2" title="Permalink to this headline">¶</a></h3>
+<p><strong class="program">llvm-profdata merge</strong> takes several profile data files
+generated by PGO instrumentation and merges them together into a single
+indexed profile data file.</p>
+</div>
+<div class="section" id="options">
+<h3>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h3>
+<dl class="option">
+<dt id="cmdoption-llvm-profdata-merge-help">
+<tt class="descname">-help</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-profdata-merge-help" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print a summary of command line options.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-profdata-merge-output">
+<tt class="descname">-output</tt><tt class="descclassname">=output</tt><tt class="descclassname">, </tt><tt class="descname">-o</tt><tt class="descclassname">=output</tt><a class="headerlink" href="#cmdoption-llvm-profdata-merge-output" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the output file name.  <em>Output</em> cannot be <tt class="docutils literal"><span class="pre">-</span></tt> as the resulting
+indexed profile data can’t be written to standard output.</p>
+</dd></dl>
+
+</div>
+</div>
+<div class="section" id="show">
+<span id="profdata-show"></span><h2>SHOW<a class="headerlink" href="#show" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="id3">
+<h3>SYNOPSIS<a class="headerlink" href="#id3" title="Permalink to this headline">¶</a></h3>
+<p><strong class="program">llvm-profdata show</strong> [<em>options</em>] [<em>filename</em>]</p>
+</div>
+<div class="section" id="id4">
+<h3>DESCRIPTION<a class="headerlink" href="#id4" title="Permalink to this headline">¶</a></h3>
+<p><strong class="program">llvm-profdata show</strong> takes a profile data file and displays the
+information about the profile counters for this file and
+for any of the specified function(s).</p>
+<p>If <em>filename</em> is omitted or is <tt class="docutils literal"><span class="pre">-</span></tt>, then <strong>llvm-profdata show</strong> reads its
+input from standard input.</p>
+</div>
+<div class="section" id="id5">
+<h3>OPTIONS<a class="headerlink" href="#id5" title="Permalink to this headline">¶</a></h3>
+<dl class="option">
+<dt id="cmdoption-llvm-profdata-show-all-functions">
+<tt class="descname">-all-functions</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-profdata-show-all-functions" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print details for every function.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-profdata-show-counts">
+<tt class="descname">-counts</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-profdata-show-counts" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print the counter values for the displayed functions.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-profdata-show-function">
+<tt class="descname">-function</tt><tt class="descclassname">=string</tt><a class="headerlink" href="#cmdoption-llvm-profdata-show-function" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print details for a function if the function’s name contains the given string.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-profdata-show-help">
+<tt class="descname">-help</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-llvm-profdata-show-help" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print a summary of command line options.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-llvm-profdata-show-output">
+<tt class="descname">-output</tt><tt class="descclassname">=output</tt><tt class="descclassname">, </tt><tt class="descname">-o</tt><tt class="descclassname">=output</tt><a class="headerlink" href="#cmdoption-llvm-profdata-show-output" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the output file name.  If <em>output</em> is <tt class="docutils literal"><span class="pre">-</span></tt> or it isn’t specified,
+then the output is sent to standard output.</p>
+</dd></dl>
+
+</div>
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-profdata</strong> returns 1 if the command is omitted or is invalid,
+if it cannot read input files, or if there is a mismatch between their data.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-stress.html" title="llvm-stress - generate random .ll files"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-cov.html" title="llvm-cov - emit coverage information"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-readobj.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-readobj.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-readobj.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-readobj.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,212 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-readobj - LLVM Object Reader — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="Getting Started with the LLVM System" href="../GettingStarted.html" />
+    <link rel="prev" title="llvm-build - LLVM Project Build Utility" href="llvm-build.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">
+      <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="../GettingStarted.html" title="Getting Started with the LLVM System"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-build.html" title="llvm-build - LLVM Project Build Utility"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-readobj-llvm-object-reader">
+<h1>llvm-readobj - LLVM Object Reader<a class="headerlink" href="#llvm-readobj-llvm-object-reader" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-readobj</strong> [<em>options</em>] [<em>input...</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong class="program">llvm-readobj</strong> tool displays low-level format-specific information
+about one or more object files. The tool and its output is primarily designed
+for use in FileCheck-based tests.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<p>If <tt class="docutils literal"><span class="pre">input</span></tt> is “<tt class="docutils literal"><span class="pre">-</span></tt>” or omitted, <strong class="program">llvm-readobj</strong> reads from standard
+input. Otherwise, it will read from the specified <tt class="docutils literal"><span class="pre">filenames</span></tt>.</p>
+<dl class="option">
+<dt id="cmdoption-help">
+<tt class="descname">-help</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-help" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print a summary of command line options.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-version">
+<tt class="descname">-version</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-version" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display the version of this program</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-file-headers">
+<tt class="descname">-file-headers</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-h</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-file-headers" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display file headers.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-sections">
+<tt class="descname">-sections</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-s</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-sections" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display all sections.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-section-data">
+<tt class="descname">-section-data</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-sd</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-section-data" title="Permalink to this definition">¶</a></dt>
+<dd><p>When used with <tt class="docutils literal"><span class="pre">-sections</span></tt>, display section data for each section shown.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-section-relocations">
+<tt class="descname">-section-relocations</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-sr</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-section-relocations" title="Permalink to this definition">¶</a></dt>
+<dd><p>When used with <tt class="docutils literal"><span class="pre">-sections</span></tt>, display relocations for each section shown.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-section-symbols">
+<tt class="descname">-section-symbols</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-st</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-section-symbols" title="Permalink to this definition">¶</a></dt>
+<dd><p>When used with <tt class="docutils literal"><span class="pre">-sections</span></tt>, display symbols for each section shown.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-relocations">
+<tt class="descname">-relocations</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-r</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-relocations" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display the relocation entries in the file.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-symbols">
+<tt class="descname">-symbols</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-t</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-symbols" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display the symbol table.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-dyn-symbols">
+<tt class="descname">-dyn-symbols</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-dyn-symbols" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display the dynamic symbol table (only for ELF object files).</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-unwind">
+<tt class="descname">-unwind</tt><tt class="descclassname"></tt><tt class="descclassname">, </tt><tt class="descname">-u</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-unwind" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display unwind information.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-expand-relocs">
+<tt class="descname">-expand-relocs</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-expand-relocs" title="Permalink to this definition">¶</a></dt>
+<dd><p>When used with <tt class="docutils literal"><span class="pre">-relocations</span></tt>, display each relocation in an expanded
+multi-line format.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-dynamic-table">
+<tt class="descname">-dynamic-table</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-dynamic-table" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display the ELF .dynamic section table (only for ELF object files).</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-needed-libs">
+<tt class="descname">-needed-libs</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-needed-libs" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display the needed libraries (only for ELF object files).</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-program-headers">
+<tt class="descname">-program-headers</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-program-headers" title="Permalink to this definition">¶</a></dt>
+<dd><p>Display the ELF program headers (only for ELF object files).</p>
+</dd></dl>
+
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-readobj</strong> returns 0.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="../GettingStarted.html" title="Getting Started with the LLVM System"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-build.html" title="llvm-build - LLVM Project Build Utility"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-stress.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-stress.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-stress.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-stress.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,136 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-stress - generate random .ll files — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-symbolizer - convert addresses into source code locations" href="llvm-symbolizer.html" />
+    <link rel="prev" title="llvm-profdata - Profile data tool" href="llvm-profdata.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">
+      <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="llvm-symbolizer.html" title="llvm-symbolizer - convert addresses into source code locations"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-profdata.html" title="llvm-profdata - Profile data tool"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-stress-generate-random-ll-files">
+<h1>llvm-stress - generate random .ll files<a class="headerlink" href="#llvm-stress-generate-random-ll-files" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-stress</strong> [-size=filesize] [-seed=initialseed] [-o=outfile]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong class="program">llvm-stress</strong> tool is used to generate random <tt class="docutils literal"><span class="pre">.ll</span></tt> files that
+can be used to test different components of LLVM.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<dl class="option">
+<dt id="cmdoption-o">
+<tt class="descname">-o</tt><tt class="descclassname"> filename</tt><a class="headerlink" href="#cmdoption-o" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the output filename.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-size">
+<tt class="descname">-size</tt><tt class="descclassname"> size</tt><a class="headerlink" href="#cmdoption-size" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the size of the generated <tt class="docutils literal"><span class="pre">.ll</span></tt> file.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-seed">
+<tt class="descname">-seed</tt><tt class="descclassname"> seed</tt><a class="headerlink" href="#cmdoption-seed" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the seed to be used for the randomly generated instructions.</p>
+</dd></dl>
+
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-stress</strong> returns 0.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-symbolizer.html" title="llvm-symbolizer - convert addresses into source code locations"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-profdata.html" title="llvm-profdata - Profile data tool"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/llvm-symbolizer.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/llvm-symbolizer.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/llvm-symbolizer.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/llvm-symbolizer.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,206 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>llvm-symbolizer - convert addresses into source code locations — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llvm-dwarfdump - print contents of DWARF sections" href="llvm-dwarfdump.html" />
+    <link rel="prev" title="llvm-stress - generate random .ll files" href="llvm-stress.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">
+      <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="llvm-dwarfdump.html" title="llvm-dwarfdump - print contents of DWARF sections"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-stress.html" title="llvm-stress - generate random .ll files"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-symbolizer-convert-addresses-into-source-code-locations">
+<h1>llvm-symbolizer - convert addresses into source code locations<a class="headerlink" href="#llvm-symbolizer-convert-addresses-into-source-code-locations" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-symbolizer</strong> [options]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-symbolizer</strong> reads object file names and addresses from standard
+input and prints corresponding source code locations to standard output.
+If object file is specified in command line, <strong class="program">llvm-symbolizer</strong> reads
+only addresses from standard input. This
+program uses debug info sections and symbol table in the object files.</p>
+</div>
+<div class="section" id="example">
+<h2>EXAMPLE<a class="headerlink" href="#example" title="Permalink to this headline">¶</a></h2>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">$</span> cat addr.txt
+<span class="go">a.out 0x4004f4</span>
+<span class="go">/tmp/b.out 0x400528</span>
+<span class="go">/tmp/c.so 0x710</span>
+<span class="go">/tmp/mach_universal_binary:i386 0x1f84</span>
+<span class="go">/tmp/mach_universal_binary:x86_64 0x100000f24</span>
+<span class="gp">$</span> llvm-symbolizer < addr.txt
+<span class="go">main</span>
+<span class="go">/tmp/a.cc:4</span>
+
+<span class="go">f(int, int)</span>
+<span class="go">/tmp/b.cc:11</span>
+
+<span class="go">h_inlined_into_g</span>
+<span class="go">/tmp/header.h:2</span>
+<span class="go">g_inlined_into_f</span>
+<span class="go">/tmp/header.h:7</span>
+<span class="go">f_inlined_into_main</span>
+<span class="go">/tmp/source.cc:3</span>
+<span class="go">main</span>
+<span class="go">/tmp/source.cc:8</span>
+
+<span class="go">_main</span>
+<span class="go">/tmp/source_i386.cc:8</span>
+
+<span class="go">_main</span>
+<span class="go">/tmp/source_x86_64.cc:8</span>
+<span class="gp">$</span> cat addr2.txt
+<span class="go">0x4004f4</span>
+<span class="go">0x401000</span>
+<span class="gp">$</span> llvm-symbolizer -obj<span class="o">=</span>a.out < addr2.txt
+<span class="go">main</span>
+<span class="go">/tmp/a.cc:4</span>
+
+<span class="go">foo(int)</span>
+<span class="go">/tmp/a.cc:12</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<dl class="option">
+<dt id="cmdoption-obj">
+<tt class="descname">-obj</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-obj" title="Permalink to this definition">¶</a></dt>
+<dd><p>Path to object file to be symbolized.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-functions">
+<tt class="descname">-functions</tt><tt class="descclassname">=[none|short|linkage]</tt><a class="headerlink" href="#cmdoption-functions" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the way function names are printed (omit function name,
+print short function name, or print full linkage name, respectively).
+Defaults to <tt class="docutils literal"><span class="pre">linkage</span></tt>.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-use-symbol-table">
+<tt class="descname">-use-symbol-table</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-use-symbol-table" title="Permalink to this definition">¶</a></dt>
+<dd><p>Prefer function names stored in symbol table to function names
+in debug info sections. Defaults to true.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-demangle">
+<tt class="descname">-demangle</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-demangle" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print demangled function names. Defaults to true.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-inlining">
+<tt class="descname">-inlining</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-inlining" title="Permalink to this definition">¶</a></dt>
+<dd><p>If a source code location is in an inlined function, prints all the
+inlnied frames. Defaults to true.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-default-arch">
+<tt class="descname">-default-arch</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-default-arch" title="Permalink to this definition">¶</a></dt>
+<dd><p>If a binary contains object files for multiple architectures (e.g. it is a
+Mach-O universal binary), symbolize the object file for a given architecture.
+You can also specify architecture by writing <tt class="docutils literal"><span class="pre">binary_name:arch_name</span></tt> in the
+input (see example above). If architecture is not specified in either way,
+address will not be symbolized. Defaults to empty string.</p>
+</dd></dl>
+
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">llvm-symbolizer</strong> returns 0. Other exit codes imply internal program error.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llvm-dwarfdump.html" title="llvm-dwarfdump - print contents of DWARF sections"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-stress.html" title="llvm-stress - generate random .ll files"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/opt.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/opt.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/opt.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/opt.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,257 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>opt - LLVM optimizer — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="llc - LLVM static compiler" href="llc.html" />
+    <link rel="prev" title="llvm-dis - LLVM disassembler" href="llvm-dis.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">
+      <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="llc.html" title="llc - LLVM static compiler"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="llvm-dis.html" title="llvm-dis - LLVM disassembler"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="opt-llvm-optimizer">
+<h1>opt - LLVM optimizer<a class="headerlink" href="#opt-llvm-optimizer" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">opt</strong> [<em>options</em>] [<em>filename</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p>The <strong class="program">opt</strong> command is the modular LLVM optimizer and analyzer.  It
+takes LLVM source files as input, runs the specified optimizations or analyses
+on it, and then outputs the optimized file or the analysis results.  The
+function of <strong class="program">opt</strong> depends on whether the <em class="xref std std-option">-analyze</em> option is
+given.</p>
+<p>When <em class="xref std std-option">-analyze</em> is specified, <strong class="program">opt</strong> performs various analyses
+of the input source.  It will usually print the results on standard output, but
+in a few cases, it will print output to standard error or generate a file with
+the analysis output, which is usually done when the output is meant for another
+program.</p>
+<p>While <em class="xref std std-option">-analyze</em> is <em>not</em> given, <strong class="program">opt</strong> attempts to produce an
+optimized output file.  The optimizations available via <strong class="program">opt</strong> depend
+upon what libraries were linked into it as well as any additional libraries
+that have been loaded with the <a class="reference internal" href="#cmdoption-load"><em class="xref std std-option">-load</em></a> option.  Use the <a class="reference internal" href="#cmdoption-help"><em class="xref std std-option">-help</em></a>
+option to determine what optimizations you can use.</p>
+<p>If <tt class="docutils literal"><span class="pre">filename</span></tt> is omitted from the command line or is “<tt class="docutils literal"><span class="pre">-</span></tt>”, <strong class="program">opt</strong>
+reads its input from standard input.  Inputs can be in either the LLVM assembly
+language format (<tt class="docutils literal"><span class="pre">.ll</span></tt>) or the LLVM bitcode format (<tt class="docutils literal"><span class="pre">.bc</span></tt>).</p>
+<p>If an output filename is not specified with the <a class="reference internal" href="#cmdoption-o"><em class="xref std std-option">-o</em></a> option,
+<strong class="program">opt</strong> writes its output to the standard output.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<dl class="option">
+<dt id="cmdoption-f">
+<tt class="descname">-f</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-f" title="Permalink to this definition">¶</a></dt>
+<dd><p>Enable binary output on terminals.  Normally, <strong class="program">opt</strong> will refuse to
+write raw bitcode output if the output stream is a terminal.  With this option,
+<strong class="program">opt</strong> will write raw bitcode regardless of the output device.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-help">
+<tt class="descname">-help</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-help" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print a summary of command line options.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-o">
+<tt class="descname">-o</tt><tt class="descclassname"> <filename></tt><a class="headerlink" href="#cmdoption-o" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the output filename.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-S">
+<tt class="descname">-S</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-S" title="Permalink to this definition">¶</a></dt>
+<dd><p>Write output in LLVM intermediate language (instead of bitcode).</p>
+</dd></dl>
+
+<dl class="option">
+<dt>
+<tt class="descname">-{passname}</tt></dt>
+<dd><p><strong class="program">opt</strong> provides the ability to run any of LLVM’s optimization or
+analysis passes in any order.  The <a class="reference internal" href="#cmdoption-help"><em class="xref std std-option">-help</em></a> option lists all the passes
+available.  The order in which the options occur on the command line are the
+order in which they are executed (within pass constraints).</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-std-compile-opts">
+<tt class="descname">-std-compile-opts</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-std-compile-opts" title="Permalink to this definition">¶</a></dt>
+<dd><p>This is short hand for a standard list of <em>compile time optimization</em> passes.
+It might be useful for other front end compilers as well.  To discover the
+full set of options available, use the following command:</p>
+<div class="highlight-sh"><div class="highlight"><pre>llvm-as < /dev/null | opt -std-compile-opts -disable-output -debug-pass<span class="o">=</span>Arguments
+</pre></div>
+</div>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-disable-inlining">
+<tt class="descname">-disable-inlining</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-disable-inlining" title="Permalink to this definition">¶</a></dt>
+<dd><p>This option is only meaningful when <a class="reference internal" href="#cmdoption-std-compile-opts"><em class="xref std std-option">-std-compile-opts</em></a> is given.  It
+simply removes the inlining pass from the standard list.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-disable-opt">
+<tt class="descname">-disable-opt</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-disable-opt" title="Permalink to this definition">¶</a></dt>
+<dd><p>This option is only meaningful when <a class="reference internal" href="#cmdoption-std-compile-opts"><em class="xref std std-option">-std-compile-opts</em></a> is given.  It
+disables most, but not all, of the <a class="reference internal" href="#cmdoption-std-compile-opts"><em class="xref std std-option">-std-compile-opts</em></a>.  The ones that
+remain are <em class="xref std std-option">-verify</em>, <em class="xref std std-option">-lower-setjmp</em>, and
+<em class="xref std std-option">-funcresolve</em>.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-strip-debug">
+<tt class="descname">-strip-debug</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-strip-debug" title="Permalink to this definition">¶</a></dt>
+<dd><p>This option causes opt to strip debug information from the module before
+applying other optimizations.  It is essentially the same as <em class="xref std std-option">-strip</em>
+but it ensures that stripping of debug information is done first.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-verify-each">
+<tt class="descname">-verify-each</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-verify-each" title="Permalink to this definition">¶</a></dt>
+<dd><p>This option causes opt to add a verify pass after every pass otherwise
+specified on the command line (including <em class="xref std std-option">-verify</em>).  This is useful
+for cases where it is suspected that a pass is creating an invalid module but
+it is not clear which pass is doing it.  The combination of
+<a class="reference internal" href="#cmdoption-std-compile-opts"><em class="xref std std-option">-std-compile-opts</em></a> and <a class="reference internal" href="#cmdoption-verify-each"><em class="xref std std-option">-verify-each</em></a> can quickly track down
+this kind of problem.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-stats">
+<tt class="descname">-stats</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-stats" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print statistics.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-time-passes">
+<tt class="descname">-time-passes</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-time-passes" title="Permalink to this definition">¶</a></dt>
+<dd><p>Record the amount of time needed for each pass and print it to standard
+error.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-debug">
+<tt class="descname">-debug</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-debug" title="Permalink to this definition">¶</a></dt>
+<dd><p>If this is a debug build, this option will enable debug printouts from passes
+which use the <tt class="docutils literal"><span class="pre">DEBUG()</span></tt> macro.  See the <a class="reference external" href="../ProgrammersManual.html">LLVM Programmer’s Manual</a>, section <tt class="docutils literal"><span class="pre">#DEBUG</span></tt> for more information.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-load">
+<tt class="descname">-load</tt><tt class="descclassname">=<plugin></tt><a class="headerlink" href="#cmdoption-load" title="Permalink to this definition">¶</a></dt>
+<dd><p>Load the dynamic object <tt class="docutils literal"><span class="pre">plugin</span></tt>.  This object should register new
+optimization or analysis passes.  Once loaded, the object will add new command
+line options to enable various optimizations or analyses.  To see the new
+complete list of optimizations, use the <a class="reference internal" href="#cmdoption-help"><em class="xref std std-option">-help</em></a> and <a class="reference internal" href="#cmdoption-load"><em class="xref std std-option">-load</em></a>
+options together.  For example:</p>
+<div class="highlight-sh"><div class="highlight"><pre>opt -load<span class="o">=</span>plugin.so -help
+</pre></div>
+</div>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-p">
+<tt class="descname">-p</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-p" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print module after each transformation.</p>
+</dd></dl>
+
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p>If <strong class="program">opt</strong> succeeds, it will exit with 0.  Otherwise, if an error
+occurs, it will exit with a non-zero value.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="llc.html" title="llc - LLVM static compiler"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="llvm-dis.html" title="llvm-dis - LLVM disassembler"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandGuide/tblgen.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandGuide/tblgen.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandGuide/tblgen.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandGuide/tblgen.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,272 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>tblgen - Target Description To C++ Code Generator — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '../',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="../index.html" />
+    <link rel="up" title="LLVM Command Guide" href="index.html" />
+    <link rel="next" title="lit - LLVM Integrated Tester" href="lit.html" />
+    <link rel="prev" title="FileCheck - Flexible pattern matching file verifier" href="FileCheck.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">
+      <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="lit.html" title="lit - LLVM Integrated Tester"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="FileCheck.html" title="FileCheck - Flexible pattern matching file verifier"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" accesskey="U">LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="tblgen-target-description-to-c-code-generator">
+<h1>tblgen - Target Description To C++ Code Generator<a class="headerlink" href="#tblgen-target-description-to-c-code-generator" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="synopsis">
+<h2>SYNOPSIS<a class="headerlink" href="#synopsis" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">tblgen</strong> [<em>options</em>] [<em>filename</em>]</p>
+</div>
+<div class="section" id="description">
+<h2>DESCRIPTION<a class="headerlink" href="#description" title="Permalink to this headline">¶</a></h2>
+<p><strong class="program">tblgen</strong> translates from target description (<tt class="docutils literal"><span class="pre">.td</span></tt>) files into C++
+code that can be included in the definition of an LLVM target library.  Most
+users of LLVM will not need to use this program.  It is only for assisting with
+writing an LLVM target backend.</p>
+<p>The input and output of <strong class="program">tblgen</strong> is beyond the scope of this short
+introduction; please see the <a class="reference internal" href="../TableGen/index.html"><em>introduction to TableGen</em></a>.</p>
+<p>The <em>filename</em> argument specifies the name of a Target Description (<tt class="docutils literal"><span class="pre">.td</span></tt>)
+file to read as input.</p>
+</div>
+<div class="section" id="options">
+<h2>OPTIONS<a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h2>
+<dl class="option">
+<dt id="cmdoption-tblgen-help">
+<tt class="descname">-help</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-help" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print a summary of command line options.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-o">
+<tt class="descname">-o</tt><tt class="descclassname"> filename</tt><a class="headerlink" href="#cmdoption-tblgen-o" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify the output file name.  If <tt class="docutils literal"><span class="pre">filename</span></tt> is <tt class="docutils literal"><span class="pre">-</span></tt>, then
+<strong class="program">tblgen</strong> sends its output to standard output.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-I">
+<tt class="descname">-I</tt><tt class="descclassname"> directory</tt><a class="headerlink" href="#cmdoption-tblgen-I" title="Permalink to this definition">¶</a></dt>
+<dd><p>Specify where to find other target description files for inclusion.  The
+<tt class="docutils literal"><span class="pre">directory</span></tt> value should be a full or partial path to a directory that
+contains target description files.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-asmparsernum">
+<tt class="descname">-asmparsernum</tt><tt class="descclassname"> N</tt><a class="headerlink" href="#cmdoption-tblgen-asmparsernum" title="Permalink to this definition">¶</a></dt>
+<dd><p>Make -gen-asm-parser emit assembly writer number <tt class="docutils literal"><span class="pre">N</span></tt>.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-asmwriternum">
+<tt class="descname">-asmwriternum</tt><tt class="descclassname"> N</tt><a class="headerlink" href="#cmdoption-tblgen-asmwriternum" title="Permalink to this definition">¶</a></dt>
+<dd><p>Make -gen-asm-writer emit assembly writer number <tt class="docutils literal"><span class="pre">N</span></tt>.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-class">
+<tt class="descname">-class</tt><tt class="descclassname"> className</tt><a class="headerlink" href="#cmdoption-tblgen-class" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print the enumeration list for this class.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-print-records">
+<tt class="descname">-print-records</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-print-records" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print all records to standard output (default).</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-print-enums">
+<tt class="descname">-print-enums</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-print-enums" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print enumeration values for a class.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-print-sets">
+<tt class="descname">-print-sets</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-print-sets" title="Permalink to this definition">¶</a></dt>
+<dd><p>Print expanded sets for testing DAG exprs.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-emitter">
+<tt class="descname">-gen-emitter</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-emitter" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate machine code emitter.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-register-info">
+<tt class="descname">-gen-register-info</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-register-info" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate registers and register classes info.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-instr-info">
+<tt class="descname">-gen-instr-info</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-instr-info" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate instruction descriptions.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-asm-writer">
+<tt class="descname">-gen-asm-writer</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-asm-writer" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate the assembly writer.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-disassembler">
+<tt class="descname">-gen-disassembler</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-disassembler" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate disassembler.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-pseudo-lowering">
+<tt class="descname">-gen-pseudo-lowering</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-pseudo-lowering" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate pseudo instruction lowering.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-dag-isel">
+<tt class="descname">-gen-dag-isel</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-dag-isel" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate a DAG (Directed Acycle Graph) instruction selector.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-asm-matcher">
+<tt class="descname">-gen-asm-matcher</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-asm-matcher" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate assembly instruction matcher.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-dfa-packetizer">
+<tt class="descname">-gen-dfa-packetizer</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-dfa-packetizer" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate DFA Packetizer for VLIW targets.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-fast-isel">
+<tt class="descname">-gen-fast-isel</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-fast-isel" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate a “fast” instruction selector.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-subtarget">
+<tt class="descname">-gen-subtarget</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-subtarget" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate subtarget enumerations.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-intrinsic">
+<tt class="descname">-gen-intrinsic</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-intrinsic" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate intrinsic information.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-tgt-intrinsic">
+<tt class="descname">-gen-tgt-intrinsic</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-tgt-intrinsic" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate target intrinsic information.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-gen-enhanced-disassembly-info">
+<tt class="descname">-gen-enhanced-disassembly-info</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-gen-enhanced-disassembly-info" title="Permalink to this definition">¶</a></dt>
+<dd><p>Generate enhanced disassembly info.</p>
+</dd></dl>
+
+<dl class="option">
+<dt id="cmdoption-tblgen-version">
+<tt class="descname">-version</tt><tt class="descclassname"></tt><a class="headerlink" href="#cmdoption-tblgen-version" title="Permalink to this definition">¶</a></dt>
+<dd><p>Show the version number of this program.</p>
+</dd></dl>
+
+</div>
+<div class="section" id="exit-status">
+<h2>EXIT STATUS<a class="headerlink" href="#exit-status" title="Permalink to this headline">¶</a></h2>
+<p>If <strong class="program">tblgen</strong> succeeds, it will exit with 0.  Otherwise, if an error
+occurs, it will exit with a non-zero value.</p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="lit.html" title="lit - LLVM Integrated Tester"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="FileCheck.html" title="FileCheck - Flexible pattern matching file verifier"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="../index.html">Documentation</a>»</li>
+
+          <li><a href="index.html" >LLVM Command Guide</a> »</li> 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CommandLine.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CommandLine.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CommandLine.html (added)
+++ www-releases/trunk/3.5.1/docs/CommandLine.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,1562 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>CommandLine 2.0 Library Manual — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="Architecture & Platform Information for Compiler Writers" href="CompilerWriterInfo.html" />
+    <link rel="prev" title="LLVM Coding Standards" href="CodingStandards.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">
+      <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="CompilerWriterInfo.html" title="Architecture & Platform Information for Compiler Writers"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="CodingStandards.html" title="LLVM Coding Standards"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="commandline-2-0-library-manual">
+<h1>CommandLine 2.0 Library Manual<a class="headerlink" href="#commandline-2-0-library-manual" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#introduction" id="id8">Introduction</a></li>
+<li><a class="reference internal" href="#quick-start-guide" id="id9">Quick Start Guide</a><ul>
+<li><a class="reference internal" href="#boolean-arguments" id="id10">Boolean Arguments</a></li>
+<li><a class="reference internal" href="#argument-aliases" id="id11">Argument Aliases</a></li>
+<li><a class="reference internal" href="#selecting-an-alternative-from-a-set-of-possibilities" id="id12">Selecting an alternative from a set of possibilities</a></li>
+<li><a class="reference internal" href="#named-alternatives" id="id13">Named Alternatives</a></li>
+<li><a class="reference internal" href="#parsing-a-list-of-options" id="id14">Parsing a list of options</a></li>
+<li><a class="reference internal" href="#collecting-options-as-a-set-of-flags" id="id15">Collecting options as a set of flags</a></li>
+<li><a class="reference internal" href="#adding-freeform-text-to-help-output" id="id16">Adding freeform text to help output</a></li>
+<li><a class="reference internal" href="#grouping-options-into-categories" id="id17">Grouping options into categories</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#reference-guide" id="id18">Reference Guide</a><ul>
+<li><a class="reference internal" href="#positional-options" id="id19">Positional Arguments</a><ul>
+<li><a class="reference internal" href="#specifying-positional-options-with-hyphens" id="id20">Specifying positional options with hyphens</a></li>
+<li><a class="reference internal" href="#determining-absolute-position-with-getposition" id="id21">Determining absolute position with getPosition()</a></li>
+<li><a class="reference internal" href="#the-cl-consumeafter-modifier" id="id22">The <tt class="docutils literal"><span class="pre">cl::ConsumeAfter</span></tt> modifier</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#internal-vs-external-storage" id="id23">Internal vs External Storage</a></li>
+<li><a class="reference internal" href="#option-attributes" id="id24">Option Attributes</a></li>
+<li><a class="reference internal" href="#option-modifiers" id="id25">Option Modifiers</a><ul>
+<li><a class="reference internal" href="#hiding-an-option-from-help-output" id="id26">Hiding an option from <tt class="docutils literal"><span class="pre">-help</span></tt> output</a></li>
+<li><a class="reference internal" href="#controlling-the-number-of-occurrences-required-and-allowed" id="id27">Controlling the number of occurrences required and allowed</a></li>
+<li><a class="reference internal" href="#controlling-whether-or-not-a-value-must-be-specified" id="id28">Controlling whether or not a value must be specified</a></li>
+<li><a class="reference internal" href="#controlling-other-formatting-options" id="id29">Controlling other formatting options</a></li>
+<li><a class="reference internal" href="#miscellaneous-option-modifiers" id="id30">Miscellaneous option modifiers</a></li>
+<li><a class="reference internal" href="#response-files" id="id31">Response files</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#top-level-classes-and-functions" id="id32">Top-Level Classes and Functions</a><ul>
+<li><a class="reference internal" href="#the-cl-getregisteredoptions-function" id="id33">The <tt class="docutils literal"><span class="pre">cl::getRegisteredOptions</span></tt> function</a></li>
+<li><a class="reference internal" href="#the-cl-parsecommandlineoptions-function" id="id34">The <tt class="docutils literal"><span class="pre">cl::ParseCommandLineOptions</span></tt> function</a></li>
+<li><a class="reference internal" href="#the-cl-parseenvironmentoptions-function" id="id35">The <tt class="docutils literal"><span class="pre">cl::ParseEnvironmentOptions</span></tt> function</a></li>
+<li><a class="reference internal" href="#the-cl-setversionprinter-function" id="id36">The <tt class="docutils literal"><span class="pre">cl::SetVersionPrinter</span></tt> function</a></li>
+<li><a class="reference internal" href="#the-cl-opt-class" id="id37">The <tt class="docutils literal"><span class="pre">cl::opt</span></tt> class</a></li>
+<li><a class="reference internal" href="#the-cl-list-class" id="id38">The <tt class="docutils literal"><span class="pre">cl::list</span></tt> class</a></li>
+<li><a class="reference internal" href="#the-cl-bits-class" id="id39">The <tt class="docutils literal"><span class="pre">cl::bits</span></tt> class</a></li>
+<li><a class="reference internal" href="#the-cl-alias-class" id="id40">The <tt class="docutils literal"><span class="pre">cl::alias</span></tt> class</a></li>
+<li><a class="reference internal" href="#the-cl-extrahelp-class" id="id41">The <tt class="docutils literal"><span class="pre">cl::extrahelp</span></tt> class</a></li>
+<li><a class="reference internal" href="#the-cl-optioncategory-class" id="id42">The <tt class="docutils literal"><span class="pre">cl::OptionCategory</span></tt> class</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#builtin-parsers" id="id43">Builtin parsers</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#extending-the-library" id="id44">Extension Guide</a><ul>
+<li><a class="reference internal" href="#writing-a-custom-parser" id="id45">Writing a custom parser</a></li>
+<li><a class="reference internal" href="#exploiting-external-storage" id="id46">Exploiting external storage</a></li>
+</ul>
+</li>
+</ul>
+</div>
+<div class="section" id="introduction">
+<h2><a class="toc-backref" href="#id8">Introduction</a><a class="headerlink" href="#introduction" title="Permalink to this headline">¶</a></h2>
+<p>This document describes the CommandLine argument processing library.  It will
+show you how to use it, and what it can do.  The CommandLine library uses a
+declarative approach to specifying the command line options that your program
+takes.  By default, these options declarations implicitly hold the value parsed
+for the option declared (of course this <a class="reference internal" href="#can-be-changed">can be changed</a>).</p>
+<p>Although there are a <strong>lot</strong> of command line argument parsing libraries out
+there in many different languages, none of them fit well with what I needed.  By
+looking at the features and problems of other libraries, I designed the
+CommandLine library to have the following features:</p>
+<ol class="arabic simple">
+<li>Speed: The CommandLine library is very quick and uses little resources.  The
+parsing time of the library is directly proportional to the number of
+arguments parsed, not the number of options recognized.  Additionally,
+command line argument values are captured transparently into user defined
+global variables, which can be accessed like any other variable (and with the
+same performance).</li>
+<li>Type Safe: As a user of CommandLine, you don’t have to worry about
+remembering the type of arguments that you want (is it an int?  a string? a
+bool? an enum?) and keep casting it around.  Not only does this help prevent
+error prone constructs, it also leads to dramatically cleaner source code.</li>
+<li>No subclasses required: To use CommandLine, you instantiate variables that
+correspond to the arguments that you would like to capture, you don’t
+subclass a parser.  This means that you don’t have to write <strong>any</strong>
+boilerplate code.</li>
+<li>Globally accessible: Libraries can specify command line arguments that are
+automatically enabled in any tool that links to the library.  This is
+possible because the application doesn’t have to keep a list of arguments to
+pass to the parser.  This also makes supporting <a class="reference internal" href="#dynamically-loaded-options">dynamically loaded options</a>
+trivial.</li>
+<li>Cleaner: CommandLine supports enum and other types directly, meaning that
+there is less error and more security built into the library.  You don’t have
+to worry about whether your integral command line argument accidentally got
+assigned a value that is not valid for your enum type.</li>
+<li>Powerful: The CommandLine library supports many different types of arguments,
+from simple <a class="reference internal" href="#boolean-flags">boolean flags</a> to <a class="reference internal" href="#scalars-arguments">scalars arguments</a> (<a class="reference internal" href="#strings">strings</a>,
+<a class="reference internal" href="#integers">integers</a>, <a class="reference internal" href="#enums">enums</a>, <a class="reference internal" href="#doubles">doubles</a>), to <a class="reference internal" href="#lists-of-arguments">lists of arguments</a>.  This is
+possible because CommandLine is...</li>
+<li>Extensible: It is very simple to add a new argument type to CommandLine.
+Simply specify the parser that you want to use with the command line option
+when you declare it. <a class="reference internal" href="#custom-parsers">Custom parsers</a> are no problem.</li>
+<li>Labor Saving: The CommandLine library cuts down on the amount of grunt work
+that you, the user, have to do.  For example, it automatically provides a
+<tt class="docutils literal"><span class="pre">-help</span></tt> option that shows the available command line options for your tool.
+Additionally, it does most of the basic correctness checking for you.</li>
+<li>Capable: The CommandLine library can handle lots of different forms of
+options often found in real programs.  For example, <a class="reference internal" href="#positional">positional</a> arguments,
+<tt class="docutils literal"><span class="pre">ls</span></tt> style <a class="reference internal" href="#grouping">grouping</a> options (to allow processing ‘<tt class="docutils literal"><span class="pre">ls</span> <span class="pre">-lad</span></tt>‘
+naturally), <tt class="docutils literal"><span class="pre">ld</span></tt> style <a class="reference internal" href="#prefix">prefix</a> options (to parse ‘<tt class="docutils literal"><span class="pre">-lmalloc</span>
+<span class="pre">-L/usr/lib</span></tt>‘), and interpreter style options.</li>
+</ol>
+<p>This document will hopefully let you jump in and start using CommandLine in your
+utility quickly and painlessly.  Additionally it should be a simple reference
+manual to figure out how stuff works.</p>
+</div>
+<div class="section" id="quick-start-guide">
+<h2><a class="toc-backref" href="#id9">Quick Start Guide</a><a class="headerlink" href="#quick-start-guide" title="Permalink to this headline">¶</a></h2>
+<p>This section of the manual runs through a simple CommandLine’ification of a
+basic compiler tool.  This is intended to show you how to jump into using the
+CommandLine library in your own program, and show you some of the cool things it
+can do.</p>
+<p>To start out, you need to include the CommandLine header file into your program:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#include "llvm/Support/CommandLine.h"</span>
+</pre></div>
+</div>
+<p>Additionally, you need to add this as the first line of your main program:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="kt">int</span> <span class="nf">main</span><span class="p">(</span><span class="kt">int</span> <span class="n">argc</span><span class="p">,</span> <span class="kt">char</span> <span class="o">**</span><span class="n">argv</span><span class="p">)</span> <span class="p">{</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">ParseCommandLineOptions</span><span class="p">(</span><span class="n">argc</span><span class="p">,</span> <span class="n">argv</span><span class="p">);</span>
+  <span class="p">...</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>... which actually parses the arguments and fills in the variable declarations.</p>
+<p>Now that you are ready to support command line arguments, we need to tell the
+system which ones we want, and what type of arguments they are.  The CommandLine
+library uses a declarative syntax to model command line arguments with the
+global variable declarations that capture the parsed values.  This means that
+for every command line option that you would like to support, there should be a
+global variable declaration to capture the result.  For example, in a compiler,
+we would like to support the Unix-standard ‘<tt class="docutils literal"><span class="pre">-o</span> <span class="pre"><filename></span></tt>‘ option to specify
+where to put the output.  With the CommandLine library, this is represented like
+this:</p>
+<div class="highlight-c++" id="here"><span id="scalars-arguments"></span><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">string</span><span class="o">></span> <span class="n">OutputFilename</span><span class="p">(</span><span class="s">"o"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Specify output filename"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">value_desc</span><span class="p">(</span><span class="s">"filename"</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>This declares a global variable “<tt class="docutils literal"><span class="pre">OutputFilename</span></tt>” that is used to capture the
+result of the “<tt class="docutils literal"><span class="pre">o</span></tt>” argument (first parameter).  We specify that this is a
+simple scalar option by using the “<tt class="docutils literal"><span class="pre">cl::opt</span></tt>” template (as opposed to the
+“<tt class="docutils literal"><span class="pre">cl::list</span></tt>” template), and tell the CommandLine library that the data
+type that we are parsing is a string.</p>
+<p>The second and third parameters (which are optional) are used to specify what to
+output for the “<tt class="docutils literal"><span class="pre">-help</span></tt>” option.  In this case, we get a line that looks like
+this:</p>
+<div class="highlight-python"><pre>USAGE: compiler [options]
+
+OPTIONS:
+  -help             - display available options (-help-hidden for more)
+  -o <filename>     - Specify output filename</pre>
+</div>
+<p>Because we specified that the command line option should parse using the
+<tt class="docutils literal"><span class="pre">string</span></tt> data type, the variable declared is automatically usable as a real
+string in all contexts that a normal C++ string object may be used.  For
+example:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="p">...</span>
+<span class="n">std</span><span class="o">::</span><span class="n">ofstream</span> <span class="n">Output</span><span class="p">(</span><span class="n">OutputFilename</span><span class="p">.</span><span class="n">c_str</span><span class="p">());</span>
+<span class="k">if</span> <span class="p">(</span><span class="n">Output</span><span class="p">.</span><span class="n">good</span><span class="p">())</span> <span class="p">...</span>
+<span class="p">...</span>
+</pre></div>
+</div>
+<p>There are many different options that you can use to customize the command line
+option handling library, but the above example shows the general interface to
+these options.  The options can be specified in any order, and are specified
+with helper functions like <a class="reference internal" href="#cl-desc">cl::desc(...)</a>, so there are no positional
+dependencies to remember.  The available options are discussed in detail in the
+<a class="reference internal" href="#reference-guide">Reference Guide</a>.</p>
+<p>Continuing the example, we would like to have our compiler take an input
+filename as well as an output filename, but we do not want the input filename to
+be specified with a hyphen (ie, not <tt class="docutils literal"><span class="pre">-filename.c</span></tt>).  To support this style of
+argument, the CommandLine library allows for <a class="reference internal" href="#positional">positional</a> arguments to be
+specified for the program.  These positional arguments are filled with command
+line parameters that are not in option form.  We use this feature like this:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">string</span><span class="o">></span> <span class="n">InputFilename</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Positional</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"<input file>"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">init</span><span class="p">(</span><span class="s">"-"</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>This declaration indicates that the first positional argument should be treated
+as the input filename.  Here we use the <a class="reference internal" href="#cl-init">cl::init</a> option to specify an initial
+value for the command line option, which is used if the option is not specified
+(if you do not specify a <a class="reference internal" href="#cl-init">cl::init</a> modifier for an option, then the default
+constructor for the data type is used to initialize the value).  Command line
+options default to being optional, so if we would like to require that the user
+always specify an input filename, we would add the <a class="reference internal" href="#cl-required">cl::Required</a> flag, and we
+could eliminate the <a class="reference internal" href="#cl-init">cl::init</a> modifier, like this:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">string</span><span class="o">></span> <span class="n">InputFilename</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Positional</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"<input file>"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">Required</span><span class="p">);</span>
+</pre></div>
+</div>
+<p>Again, the CommandLine library does not require the options to be specified in
+any particular order, so the above declaration is equivalent to:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">string</span><span class="o">></span> <span class="n">InputFilename</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Positional</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">Required</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"<input file>"</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>By simply adding the <a class="reference internal" href="#cl-required">cl::Required</a> flag, the CommandLine library will
+automatically issue an error if the argument is not specified, which shifts all
+of the command line option verification code out of your application into the
+library.  This is just one example of how using flags can alter the default
+behaviour of the library, on a per-option basis.  By adding one of the
+declarations above, the <tt class="docutils literal"><span class="pre">-help</span></tt> option synopsis is now extended to:</p>
+<div class="highlight-python"><pre>USAGE: compiler [options] <input file>
+
+OPTIONS:
+  -help             - display available options (-help-hidden for more)
+  -o <filename>     - Specify output filename</pre>
+</div>
+<p>... indicating that an input filename is expected.</p>
+<div class="section" id="boolean-arguments">
+<h3><a class="toc-backref" href="#id10">Boolean Arguments</a><a class="headerlink" href="#boolean-arguments" title="Permalink to this headline">¶</a></h3>
+<p>In addition to input and output filenames, we would like the compiler example to
+support three boolean flags: “<tt class="docutils literal"><span class="pre">-f</span></tt>” to force writing binary output to a
+terminal, “<tt class="docutils literal"><span class="pre">--quiet</span></tt>” to enable quiet mode, and “<tt class="docutils literal"><span class="pre">-q</span></tt>” for backwards
+compatibility with some of our users.  We can support these by declaring options
+of boolean type like this:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span> <span class="n">Force</span> <span class="p">(</span><span class="s">"f"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Enable binary output on terminals"</span><span class="p">));</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span> <span class="n">Quiet</span> <span class="p">(</span><span class="s">"quiet"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Don't print informational messages"</span><span class="p">));</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span> <span class="n">Quiet2</span><span class="p">(</span><span class="s">"q"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Don't print informational messages"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">Hidden</span><span class="p">);</span>
+</pre></div>
+</div>
+<p>This does what you would expect: it declares three boolean variables
+(“<tt class="docutils literal"><span class="pre">Force</span></tt>”, “<tt class="docutils literal"><span class="pre">Quiet</span></tt>”, and “<tt class="docutils literal"><span class="pre">Quiet2</span></tt>”) to recognize these options.  Note
+that the “<tt class="docutils literal"><span class="pre">-q</span></tt>” option is specified with the “<a class="reference internal" href="#cl-hidden">cl::Hidden</a>” flag.  This
+modifier prevents it from being shown by the standard “<tt class="docutils literal"><span class="pre">-help</span></tt>” output (note
+that it is still shown in the “<tt class="docutils literal"><span class="pre">-help-hidden</span></tt>” output).</p>
+<p>The CommandLine library uses a <a class="reference internal" href="#different-parser">different parser</a> for different data types.
+For example, in the string case, the argument passed to the option is copied
+literally into the content of the string variable... we obviously cannot do that
+in the boolean case, however, so we must use a smarter parser.  In the case of
+the boolean parser, it allows no options (in which case it assigns the value of
+true to the variable), or it allows the values “<tt class="docutils literal"><span class="pre">true</span></tt>” or “<tt class="docutils literal"><span class="pre">false</span></tt>” to be
+specified, allowing any of the following inputs:</p>
+<div class="highlight-python"><div class="highlight"><pre><span class="n">compiler</span> <span class="o">-</span><span class="n">f</span>          <span class="c"># No value, 'Force' == true</span>
+<span class="n">compiler</span> <span class="o">-</span><span class="n">f</span><span class="o">=</span><span class="n">true</span>     <span class="c"># Value specified, 'Force' == true</span>
+<span class="n">compiler</span> <span class="o">-</span><span class="n">f</span><span class="o">=</span><span class="n">TRUE</span>     <span class="c"># Value specified, 'Force' == true</span>
+<span class="n">compiler</span> <span class="o">-</span><span class="n">f</span><span class="o">=</span><span class="n">FALSE</span>    <span class="c"># Value specified, 'Force' == false</span>
+</pre></div>
+</div>
+<p>... you get the idea.  The <a class="reference internal" href="#bool-parser">bool parser</a> just turns the string values into
+boolean values, and rejects things like ‘<tt class="docutils literal"><span class="pre">compiler</span> <span class="pre">-f=foo</span></tt>‘.  Similarly, the
+<a class="reference internal" href="#float">float</a>, <a class="reference internal" href="#double">double</a>, and <a class="reference internal" href="#int">int</a> parsers work like you would expect, using the
+‘<tt class="docutils literal"><span class="pre">strtol</span></tt>‘ and ‘<tt class="docutils literal"><span class="pre">strtod</span></tt>‘ C library calls to parse the string value into the
+specified data type.</p>
+<p>With the declarations above, “<tt class="docutils literal"><span class="pre">compiler</span> <span class="pre">-help</span></tt>” emits this:</p>
+<div class="highlight-python"><pre>USAGE: compiler [options] <input file>
+
+OPTIONS:
+  -f     - Enable binary output on terminals
+  -o     - Override output filename
+  -quiet - Don't print informational messages
+  -help  - display available options (-help-hidden for more)</pre>
+</div>
+<p>and “<tt class="docutils literal"><span class="pre">compiler</span> <span class="pre">-help-hidden</span></tt>” prints this:</p>
+<div class="highlight-python"><pre>USAGE: compiler [options] <input file>
+
+OPTIONS:
+  -f     - Enable binary output on terminals
+  -o     - Override output filename
+  -q     - Don't print informational messages
+  -quiet - Don't print informational messages
+  -help  - display available options (-help-hidden for more)</pre>
+</div>
+<p>This brief example has shown you how to use the ‘<a class="reference internal" href="#cl-opt">cl::opt</a>‘ class to parse
+simple scalar command line arguments.  In addition to simple scalar arguments,
+the CommandLine library also provides primitives to support CommandLine option
+<a class="reference internal" href="#aliases">aliases</a>, and <a class="reference internal" href="#lists">lists</a> of options.</p>
+</div>
+<div class="section" id="argument-aliases">
+<span id="aliases"></span><h3><a class="toc-backref" href="#id11">Argument Aliases</a><a class="headerlink" href="#argument-aliases" title="Permalink to this headline">¶</a></h3>
+<p>So far, the example works well, except for the fact that we need to check the
+quiet condition like this now:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="p">...</span>
+  <span class="k">if</span> <span class="p">(</span><span class="o">!</span><span class="n">Quiet</span> <span class="o">&&</span> <span class="o">!</span><span class="n">Quiet2</span><span class="p">)</span> <span class="n">printInformationalMessage</span><span class="p">(...);</span>
+<span class="p">...</span>
+</pre></div>
+</div>
+<p>... which is a real pain!  Instead of defining two values for the same
+condition, we can use the “<a class="reference internal" href="#cl-alias">cl::alias</a>” class to make the “<tt class="docutils literal"><span class="pre">-q</span></tt>” option an
+<strong>alias</strong> for the “<tt class="docutils literal"><span class="pre">-quiet</span></tt>” option, instead of providing a value itself:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span> <span class="n">Force</span> <span class="p">(</span><span class="s">"f"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Overwrite output files"</span><span class="p">));</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span> <span class="n">Quiet</span> <span class="p">(</span><span class="s">"quiet"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Don't print informational messages"</span><span class="p">));</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">alias</span>     <span class="n">QuietA</span><span class="p">(</span><span class="s">"q"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Alias for -quiet"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">aliasopt</span><span class="p">(</span><span class="n">Quiet</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>The third line (which is the only one we modified from above) defines a “<tt class="docutils literal"><span class="pre">-q</span></tt>”
+alias that updates the “<tt class="docutils literal"><span class="pre">Quiet</span></tt>” variable (as specified by the <a class="reference internal" href="#cl-aliasopt">cl::aliasopt</a>
+modifier) whenever it is specified.  Because aliases do not hold state, the only
+thing the program has to query is the <tt class="docutils literal"><span class="pre">Quiet</span></tt> variable now.  Another nice
+feature of aliases is that they automatically hide themselves from the <tt class="docutils literal"><span class="pre">-help</span></tt>
+output (although, again, they are still visible in the <tt class="docutils literal"><span class="pre">-help-hidden</span> <span class="pre">output</span></tt>).</p>
+<p>Now the application code can simply use:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="p">...</span>
+  <span class="k">if</span> <span class="p">(</span><span class="o">!</span><span class="n">Quiet</span><span class="p">)</span> <span class="n">printInformationalMessage</span><span class="p">(...);</span>
+<span class="p">...</span>
+</pre></div>
+</div>
+<p>... which is much nicer!  The “<a class="reference internal" href="#cl-alias">cl::alias</a>” can be used to specify an
+alternative name for any variable type, and has many uses.</p>
+</div>
+<div class="section" id="selecting-an-alternative-from-a-set-of-possibilities">
+<span id="unnamed-alternatives-using-the-generic-parser"></span><h3><a class="toc-backref" href="#id12">Selecting an alternative from a set of possibilities</a><a class="headerlink" href="#selecting-an-alternative-from-a-set-of-possibilities" title="Permalink to this headline">¶</a></h3>
+<p>So far we have seen how the CommandLine library handles builtin types like
+<tt class="docutils literal"><span class="pre">std::string</span></tt>, <tt class="docutils literal"><span class="pre">bool</span></tt> and <tt class="docutils literal"><span class="pre">int</span></tt>, but how does it handle things it doesn’t
+know about, like enums or ‘<tt class="docutils literal"><span class="pre">int*</span></tt>‘s?</p>
+<p>The answer is that it uses a table-driven generic parser (unless you specify
+your own parser, as described in the <a class="reference internal" href="#extension-guide">Extension Guide</a>).  This parser maps
+literal strings to whatever type is required, and requires you to tell it what
+this mapping should be.</p>
+<p>Let’s say that we would like to add four optimization levels to our optimizer,
+using the standard flags “<tt class="docutils literal"><span class="pre">-g</span></tt>”, “<tt class="docutils literal"><span class="pre">-O0</span></tt>”, “<tt class="docutils literal"><span class="pre">-O1</span></tt>”, and “<tt class="docutils literal"><span class="pre">-O2</span></tt>”.  We
+could easily implement this with boolean options like above, but there are
+several problems with this strategy:</p>
+<ol class="arabic simple">
+<li>A user could specify more than one of the options at a time, for example,
+“<tt class="docutils literal"><span class="pre">compiler</span> <span class="pre">-O3</span> <span class="pre">-O2</span></tt>”.  The CommandLine library would not be able to catch
+this erroneous input for us.</li>
+<li>We would have to test 4 different variables to see which ones are set.</li>
+<li>This doesn’t map to the numeric levels that we want... so we cannot easily
+see if some level >= “<tt class="docutils literal"><span class="pre">-O1</span></tt>” is enabled.</li>
+</ol>
+<p>To cope with these problems, we can use an enum value, and have the CommandLine
+library fill it in with the appropriate level directly, which is used like this:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">enum</span> <span class="n">OptLevel</span> <span class="p">{</span>
+  <span class="n">g</span><span class="p">,</span> <span class="n">O1</span><span class="p">,</span> <span class="n">O2</span><span class="p">,</span> <span class="n">O3</span>
+<span class="p">};</span>
+
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">OptLevel</span><span class="o">></span> <span class="n">OptimizationLevel</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Choose optimization level:"</span><span class="p">),</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">values</span><span class="p">(</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">g</span> <span class="p">,</span> <span class="s">"No optimizations, enable debugging"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">O1</span><span class="p">,</span> <span class="s">"Enable trivial optimizations"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">O2</span><span class="p">,</span> <span class="s">"Enable default optimizations"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">O3</span><span class="p">,</span> <span class="s">"Enable expensive optimizations"</span><span class="p">),</span>
+   <span class="n">clEnumValEnd</span><span class="p">));</span>
+
+<span class="p">...</span>
+  <span class="k">if</span> <span class="p">(</span><span class="n">OptimizationLevel</span> <span class="o">>=</span> <span class="n">O2</span><span class="p">)</span> <span class="n">doPartialRedundancyElimination</span><span class="p">(...);</span>
+<span class="p">...</span>
+</pre></div>
+</div>
+<p>This declaration defines a variable “<tt class="docutils literal"><span class="pre">OptimizationLevel</span></tt>” of the
+“<tt class="docutils literal"><span class="pre">OptLevel</span></tt>” enum type.  This variable can be assigned any of the values that
+are listed in the declaration (Note that the declaration list must be terminated
+with the “<tt class="docutils literal"><span class="pre">clEnumValEnd</span></tt>” argument!).  The CommandLine library enforces that
+the user can only specify one of the options, and it ensure that only valid enum
+values can be specified.  The “<tt class="docutils literal"><span class="pre">clEnumVal</span></tt>” macros ensure that the command
+line arguments matched the enum values.  With this option added, our help output
+now is:</p>
+<div class="highlight-python"><pre>USAGE: compiler [options] <input file>
+
+OPTIONS:
+  Choose optimization level:
+    -g          - No optimizations, enable debugging
+    -O1         - Enable trivial optimizations
+    -O2         - Enable default optimizations
+    -O3         - Enable expensive optimizations
+  -f            - Enable binary output on terminals
+  -help         - display available options (-help-hidden for more)
+  -o <filename> - Specify output filename
+  -quiet        - Don't print informational messages</pre>
+</div>
+<p>In this case, it is sort of awkward that flag names correspond directly to enum
+names, because we probably don’t want a enum definition named “<tt class="docutils literal"><span class="pre">g</span></tt>” in our
+program.  Because of this, we can alternatively write this example like this:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">enum</span> <span class="n">OptLevel</span> <span class="p">{</span>
+  <span class="n">Debug</span><span class="p">,</span> <span class="n">O1</span><span class="p">,</span> <span class="n">O2</span><span class="p">,</span> <span class="n">O3</span>
+<span class="p">};</span>
+
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">OptLevel</span><span class="o">></span> <span class="n">OptimizationLevel</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Choose optimization level:"</span><span class="p">),</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">values</span><span class="p">(</span>
+   <span class="n">clEnumValN</span><span class="p">(</span><span class="n">Debug</span><span class="p">,</span> <span class="s">"g"</span><span class="p">,</span> <span class="s">"No optimizations, enable debugging"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">O1</span>        <span class="p">,</span> <span class="s">"Enable trivial optimizations"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">O2</span>        <span class="p">,</span> <span class="s">"Enable default optimizations"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">O3</span>        <span class="p">,</span> <span class="s">"Enable expensive optimizations"</span><span class="p">),</span>
+   <span class="n">clEnumValEnd</span><span class="p">));</span>
+
+<span class="p">...</span>
+  <span class="k">if</span> <span class="p">(</span><span class="n">OptimizationLevel</span> <span class="o">==</span> <span class="n">Debug</span><span class="p">)</span> <span class="n">outputDebugInfo</span><span class="p">(...);</span>
+<span class="p">...</span>
+</pre></div>
+</div>
+<p>By using the “<tt class="docutils literal"><span class="pre">clEnumValN</span></tt>” macro instead of “<tt class="docutils literal"><span class="pre">clEnumVal</span></tt>”, we can directly
+specify the name that the flag should get.  In general a direct mapping is nice,
+but sometimes you can’t or don’t want to preserve the mapping, which is when you
+would use it.</p>
+</div>
+<div class="section" id="named-alternatives">
+<h3><a class="toc-backref" href="#id13">Named Alternatives</a><a class="headerlink" href="#named-alternatives" title="Permalink to this headline">¶</a></h3>
+<p>Another useful argument form is a named alternative style.  We shall use this
+style in our compiler to specify different debug levels that can be used.
+Instead of each debug level being its own switch, we want to support the
+following options, of which only one can be specified at a time:
+“<tt class="docutils literal"><span class="pre">--debug-level=none</span></tt>”, “<tt class="docutils literal"><span class="pre">--debug-level=quick</span></tt>”,
+“<tt class="docutils literal"><span class="pre">--debug-level=detailed</span></tt>”.  To do this, we use the exact same format as our
+optimization level flags, but we also specify an option name.  For this case,
+the code looks like this:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">enum</span> <span class="n">DebugLev</span> <span class="p">{</span>
+  <span class="n">nodebuginfo</span><span class="p">,</span> <span class="n">quick</span><span class="p">,</span> <span class="n">detailed</span>
+<span class="p">};</span>
+
+<span class="c1">// Enable Debug Options to be specified on the command line</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">DebugLev</span><span class="o">></span> <span class="n">DebugLevel</span><span class="p">(</span><span class="s">"debug_level"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Set the debugging level:"</span><span class="p">),</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">values</span><span class="p">(</span>
+    <span class="n">clEnumValN</span><span class="p">(</span><span class="n">nodebuginfo</span><span class="p">,</span> <span class="s">"none"</span><span class="p">,</span> <span class="s">"disable debug information"</span><span class="p">),</span>
+     <span class="n">clEnumVal</span><span class="p">(</span><span class="n">quick</span><span class="p">,</span>               <span class="s">"enable quick debug information"</span><span class="p">),</span>
+     <span class="n">clEnumVal</span><span class="p">(</span><span class="n">detailed</span><span class="p">,</span>            <span class="s">"enable detailed debug information"</span><span class="p">),</span>
+    <span class="n">clEnumValEnd</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>This definition defines an enumerated command line variable of type “<tt class="docutils literal"><span class="pre">enum</span>
+<span class="pre">DebugLev</span></tt>”, which works exactly the same way as before.  The difference here is
+just the interface exposed to the user of your program and the help output by
+the “<tt class="docutils literal"><span class="pre">-help</span></tt>” option:</p>
+<div class="highlight-python"><pre>USAGE: compiler [options] <input file>
+
+OPTIONS:
+  Choose optimization level:
+    -g          - No optimizations, enable debugging
+    -O1         - Enable trivial optimizations
+    -O2         - Enable default optimizations
+    -O3         - Enable expensive optimizations
+  -debug_level  - Set the debugging level:
+    =none       - disable debug information
+    =quick      - enable quick debug information
+    =detailed   - enable detailed debug information
+  -f            - Enable binary output on terminals
+  -help         - display available options (-help-hidden for more)
+  -o <filename> - Specify output filename
+  -quiet        - Don't print informational messages</pre>
+</div>
+<p>Again, the only structural difference between the debug level declaration and
+the optimization level declaration is that the debug level declaration includes
+an option name (<tt class="docutils literal"><span class="pre">"debug_level"</span></tt>), which automatically changes how the library
+processes the argument.  The CommandLine library supports both forms so that you
+can choose the form most appropriate for your application.</p>
+</div>
+<div class="section" id="parsing-a-list-of-options">
+<span id="lists"></span><h3><a class="toc-backref" href="#id14">Parsing a list of options</a><a class="headerlink" href="#parsing-a-list-of-options" title="Permalink to this headline">¶</a></h3>
+<p>Now that we have the standard run-of-the-mill argument types out of the way,
+lets get a little wild and crazy.  Lets say that we want our optimizer to accept
+a <strong>list</strong> of optimizations to perform, allowing duplicates.  For example, we
+might want to run: “<tt class="docutils literal"><span class="pre">compiler</span> <span class="pre">-dce</span> <span class="pre">-constprop</span> <span class="pre">-inline</span> <span class="pre">-dce</span> <span class="pre">-strip</span></tt>”.  In this
+case, the order of the arguments and the number of appearances is very
+important.  This is what the “<tt class="docutils literal"><span class="pre">cl::list</span></tt>” template is for.  First, start by
+defining an enum of the optimizations that you would like to perform:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">enum</span> <span class="n">Opts</span> <span class="p">{</span>
+  <span class="c1">// 'inline' is a C++ keyword, so name it 'inlining'</span>
+  <span class="n">dce</span><span class="p">,</span> <span class="n">constprop</span><span class="p">,</span> <span class="n">inlining</span><span class="p">,</span> <span class="n">strip</span>
+<span class="p">};</span>
+</pre></div>
+</div>
+<p>Then define your “<tt class="docutils literal"><span class="pre">cl::list</span></tt>” variable:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">list</span><span class="o"><</span><span class="n">Opts</span><span class="o">></span> <span class="n">OptimizationList</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Available Optimizations:"</span><span class="p">),</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">values</span><span class="p">(</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">dce</span>               <span class="p">,</span> <span class="s">"Dead Code Elimination"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">constprop</span>         <span class="p">,</span> <span class="s">"Constant Propagation"</span><span class="p">),</span>
+   <span class="n">clEnumValN</span><span class="p">(</span><span class="n">inlining</span><span class="p">,</span> <span class="s">"inline"</span><span class="p">,</span> <span class="s">"Procedure Integration"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">strip</span>             <span class="p">,</span> <span class="s">"Strip Symbols"</span><span class="p">),</span>
+  <span class="n">clEnumValEnd</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>This defines a variable that is conceptually of the type
+“<tt class="docutils literal"><span class="pre">std::vector<enum</span> <span class="pre">Opts></span></tt>”.  Thus, you can access it with standard vector
+methods:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">for</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="n">i</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span> <span class="n">i</span> <span class="o">!=</span> <span class="n">OptimizationList</span><span class="p">.</span><span class="n">size</span><span class="p">();</span> <span class="o">++</span><span class="n">i</span><span class="p">)</span>
+  <span class="k">switch</span> <span class="p">(</span><span class="n">OptimizationList</span><span class="p">[</span><span class="n">i</span><span class="p">])</span>
+     <span class="p">...</span>
+</pre></div>
+</div>
+<p>... to iterate through the list of options specified.</p>
+<p>Note that the “<tt class="docutils literal"><span class="pre">cl::list</span></tt>” template is completely general and may be used with
+any data types or other arguments that you can use with the “<tt class="docutils literal"><span class="pre">cl::opt</span></tt>”
+template.  One especially useful way to use a list is to capture all of the
+positional arguments together if there may be more than one specified.  In the
+case of a linker, for example, the linker takes several ‘<tt class="docutils literal"><span class="pre">.o</span></tt>‘ files, and
+needs to capture them into a list.  This is naturally specified as:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="p">...</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">list</span><span class="o"><</span><span class="n">std</span><span class="o">::</span><span class="n">string</span><span class="o">></span> <span class="n">InputFilenames</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Positional</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"<Input files>"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">OneOrMore</span><span class="p">);</span>
+<span class="p">...</span>
+</pre></div>
+</div>
+<p>This variable works just like a “<tt class="docutils literal"><span class="pre">vector<string></span></tt>” object.  As such, accessing
+the list is simple, just like above.  In this example, we used the
+<a class="reference internal" href="#cl-oneormore">cl::OneOrMore</a> modifier to inform the CommandLine library that it is an error
+if the user does not specify any <tt class="docutils literal"><span class="pre">.o</span></tt> files on our command line.  Again, this
+just reduces the amount of checking we have to do.</p>
+</div>
+<div class="section" id="collecting-options-as-a-set-of-flags">
+<h3><a class="toc-backref" href="#id15">Collecting options as a set of flags</a><a class="headerlink" href="#collecting-options-as-a-set-of-flags" title="Permalink to this headline">¶</a></h3>
+<p>Instead of collecting sets of options in a list, it is also possible to gather
+information for enum values in a <strong>bit vector</strong>.  The representation used by the
+<a class="reference internal" href="#cl-bits">cl::bits</a> class is an <tt class="docutils literal"><span class="pre">unsigned</span></tt> integer.  An enum value is represented by a
+0/1 in the enum’s ordinal value bit position. 1 indicating that the enum was
+specified, 0 otherwise.  As each specified value is parsed, the resulting enum’s
+bit is set in the option’s bit vector:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">bits</span> <span class="o">|=</span> <span class="mi">1</span> <span class="o"><<</span> <span class="p">(</span><span class="kt">unsigned</span><span class="p">)</span><span class="k">enum</span><span class="p">;</span>
+</pre></div>
+</div>
+<p>Options that are specified multiple times are redundant.  Any instances after
+the first are discarded.</p>
+<p>Reworking the above list example, we could replace <a class="reference internal" href="#cl-list">cl::list</a> with <a class="reference internal" href="#cl-bits">cl::bits</a>:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">bits</span><span class="o"><</span><span class="n">Opts</span><span class="o">></span> <span class="n">OptimizationBits</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Available Optimizations:"</span><span class="p">),</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">values</span><span class="p">(</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">dce</span>               <span class="p">,</span> <span class="s">"Dead Code Elimination"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">constprop</span>         <span class="p">,</span> <span class="s">"Constant Propagation"</span><span class="p">),</span>
+   <span class="n">clEnumValN</span><span class="p">(</span><span class="n">inlining</span><span class="p">,</span> <span class="s">"inline"</span><span class="p">,</span> <span class="s">"Procedure Integration"</span><span class="p">),</span>
+    <span class="n">clEnumVal</span><span class="p">(</span><span class="n">strip</span>             <span class="p">,</span> <span class="s">"Strip Symbols"</span><span class="p">),</span>
+  <span class="n">clEnumValEnd</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>To test to see if <tt class="docutils literal"><span class="pre">constprop</span></tt> was specified, we can use the <tt class="docutils literal"><span class="pre">cl:bits::isSet</span></tt>
+function:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">if</span> <span class="p">(</span><span class="n">OptimizationBits</span><span class="p">.</span><span class="n">isSet</span><span class="p">(</span><span class="n">constprop</span><span class="p">))</span> <span class="p">{</span>
+  <span class="p">...</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>It’s also possible to get the raw bit vector using the <tt class="docutils literal"><span class="pre">cl::bits::getBits</span></tt>
+function:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="kt">unsigned</span> <span class="n">bits</span> <span class="o">=</span> <span class="n">OptimizationBits</span><span class="p">.</span><span class="n">getBits</span><span class="p">();</span>
+</pre></div>
+</div>
+<p>Finally, if external storage is used, then the location specified must be of
+<strong>type</strong> <tt class="docutils literal"><span class="pre">unsigned</span></tt>. In all other ways a <a class="reference internal" href="#cl-bits">cl::bits</a> option is equivalent to a
+<a class="reference internal" href="#cl-list">cl::list</a> option.</p>
+</div>
+<div class="section" id="adding-freeform-text-to-help-output">
+<span id="additional-extra-text"></span><h3><a class="toc-backref" href="#id16">Adding freeform text to help output</a><a class="headerlink" href="#adding-freeform-text-to-help-output" title="Permalink to this headline">¶</a></h3>
+<p>As our program grows and becomes more mature, we may decide to put summary
+information about what it does into the help output.  The help output is styled
+to look similar to a Unix <tt class="docutils literal"><span class="pre">man</span></tt> page, providing concise information about a
+program.  Unix <tt class="docutils literal"><span class="pre">man</span></tt> pages, however often have a description about what the
+program does.  To add this to your CommandLine program, simply pass a third
+argument to the <a class="reference internal" href="#cl-parsecommandlineoptions">cl::ParseCommandLineOptions</a> call in main.  This additional
+argument is then printed as the overview information for your program, allowing
+you to include any additional information that you want.  For example:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="kt">int</span> <span class="nf">main</span><span class="p">(</span><span class="kt">int</span> <span class="n">argc</span><span class="p">,</span> <span class="kt">char</span> <span class="o">**</span><span class="n">argv</span><span class="p">)</span> <span class="p">{</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">ParseCommandLineOptions</span><span class="p">(</span><span class="n">argc</span><span class="p">,</span> <span class="n">argv</span><span class="p">,</span> <span class="s">" CommandLine compiler example</span><span class="se">\n\n</span><span class="s">"</span>
+                              <span class="s">"  This program blah blah blah...</span><span class="se">\n</span><span class="s">"</span><span class="p">);</span>
+  <span class="p">...</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>would yield the help output:</p>
+<div class="highlight-python"><pre>**OVERVIEW: CommandLine compiler example
+
+  This program blah blah blah...**
+
+USAGE: compiler [options] <input file>
+
+OPTIONS:
+  ...
+  -help             - display available options (-help-hidden for more)
+  -o <filename>     - Specify output filename</pre>
+</div>
+</div>
+<div class="section" id="grouping-options-into-categories">
+<span id="id1"></span><h3><a class="toc-backref" href="#id17">Grouping options into categories</a><a class="headerlink" href="#grouping-options-into-categories" title="Permalink to this headline">¶</a></h3>
+<p>If our program has a large number of options it may become difficult for users
+of our tool to navigate the output of <tt class="docutils literal"><span class="pre">-help</span></tt>. To alleviate this problem we
+can put our options into categories. This can be done by declaring option
+categories (<a class="reference internal" href="#cl-optioncategory">cl::OptionCategory</a> objects) and then placing our options into
+these categories using the <a class="reference internal" href="#cl-cat">cl::cat</a> option attribute. For example:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">OptionCategory</span> <span class="n">StageSelectionCat</span><span class="p">(</span><span class="s">"Stage Selection Options"</span><span class="p">,</span>
+                                     <span class="s">"These control which stages are run."</span><span class="p">);</span>
+
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span> <span class="n">Preprocessor</span><span class="p">(</span><span class="s">"E"</span><span class="p">,</span><span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Run preprocessor stage."</span><span class="p">),</span>
+                           <span class="n">cl</span><span class="o">::</span><span class="n">cat</span><span class="p">(</span><span class="n">StageSelectionCat</span><span class="p">));</span>
+
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span> <span class="n">NoLink</span><span class="p">(</span><span class="s">"c"</span><span class="p">,</span><span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Run all stages except linking."</span><span class="p">),</span>
+                     <span class="n">cl</span><span class="o">::</span><span class="n">cat</span><span class="p">(</span><span class="n">StageSelectionCat</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>The output of <tt class="docutils literal"><span class="pre">-help</span></tt> will become categorized if an option category is
+declared. The output looks something like</p>
+<div class="highlight-python"><pre>OVERVIEW: This is a small program to demo the LLVM CommandLine API
+USAGE: Sample [options]
+
+OPTIONS:
+
+  General options:
+
+    -help              - Display available options (-help-hidden for more)
+    -help-list         - Display list of available options (-help-list-hidden for more)
+
+
+  Stage Selection Options:
+  These control which stages are run.
+
+    -E                 - Run preprocessor stage.
+    -c                 - Run all stages except linking.</pre>
+</div>
+<p>In addition to the behaviour of <tt class="docutils literal"><span class="pre">-help</span></tt> changing when an option category is
+declared, the command line option <tt class="docutils literal"><span class="pre">-help-list</span></tt> becomes visible which will
+print the command line options as uncategorized list.</p>
+<p>Note that Options that are not explicitly categorized will be placed in the
+<tt class="docutils literal"><span class="pre">cl::GeneralCategory</span></tt> category.</p>
+</div>
+</div>
+<div class="section" id="reference-guide">
+<span id="id2"></span><h2><a class="toc-backref" href="#id18">Reference Guide</a><a class="headerlink" href="#reference-guide" title="Permalink to this headline">¶</a></h2>
+<p>Now that you know the basics of how to use the CommandLine library, this section
+will give you the detailed information you need to tune how command line options
+work, as well as information on more “advanced” command line option processing
+capabilities.</p>
+<div class="section" id="positional-options">
+<span id="positional-arguments-section"></span><span id="positional-arguments"></span><span id="positional-argument"></span><span id="positional"></span><span id="id3"></span><h3><a class="toc-backref" href="#id19">Positional Arguments</a><a class="headerlink" href="#positional-options" title="Permalink to this headline">¶</a></h3>
+<p>Positional arguments are those arguments that are not named, and are not
+specified with a hyphen.  Positional arguments should be used when an option is
+specified by its position alone.  For example, the standard Unix <tt class="docutils literal"><span class="pre">grep</span></tt> tool
+takes a regular expression argument, and an optional filename to search through
+(which defaults to standard input if a filename is not specified).  Using the
+CommandLine library, this would be specified as:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">string</span><span class="o">></span> <span class="n">Regex</span>   <span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Positional</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"<regular expression>"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">Required</span><span class="p">);</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">string</span><span class="o">></span> <span class="n">Filename</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Positional</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"<input file>"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">init</span><span class="p">(</span><span class="s">"-"</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>Given these two option declarations, the <tt class="docutils literal"><span class="pre">-help</span></tt> output for our grep
+replacement would look like this:</p>
+<div class="highlight-python"><pre>USAGE: spiffygrep [options] <regular expression> <input file>
+
+OPTIONS:
+  -help - display available options (-help-hidden for more)</pre>
+</div>
+<p>... and the resultant program could be used just like the standard <tt class="docutils literal"><span class="pre">grep</span></tt>
+tool.</p>
+<p>Positional arguments are sorted by their order of construction.  This means that
+command line options will be ordered according to how they are listed in a .cpp
+file, but will not have an ordering defined if the positional arguments are
+defined in multiple .cpp files.  The fix for this problem is simply to define
+all of your positional arguments in one .cpp file.</p>
+<div class="section" id="specifying-positional-options-with-hyphens">
+<h4><a class="toc-backref" href="#id20">Specifying positional options with hyphens</a><a class="headerlink" href="#specifying-positional-options-with-hyphens" title="Permalink to this headline">¶</a></h4>
+<p>Sometimes you may want to specify a value to your positional argument that
+starts with a hyphen (for example, searching for ‘<tt class="docutils literal"><span class="pre">-foo</span></tt>‘ in a file).  At
+first, you will have trouble doing this, because it will try to find an argument
+named ‘<tt class="docutils literal"><span class="pre">-foo</span></tt>‘, and will fail (and single quotes will not save you).  Note
+that the system <tt class="docutils literal"><span class="pre">grep</span></tt> has the same problem:</p>
+<div class="highlight-python"><pre>$ spiffygrep '-foo' test.txt
+Unknown command line argument '-foo'.  Try: spiffygrep -help'
+
+$ grep '-foo' test.txt
+grep: illegal option -- f
+grep: illegal option -- o
+grep: illegal option -- o
+Usage: grep -hblcnsviw pattern file . . .</pre>
+</div>
+<p>The solution for this problem is the same for both your tool and the system
+version: use the ‘<tt class="docutils literal"><span class="pre">--</span></tt>‘ marker.  When the user specifies ‘<tt class="docutils literal"><span class="pre">--</span></tt>‘ on the
+command line, it is telling the program that all options after the ‘<tt class="docutils literal"><span class="pre">--</span></tt>‘
+should be treated as positional arguments, not options.  Thus, we can use it
+like this:</p>
+<div class="highlight-python"><pre>$ spiffygrep -- -foo test.txt
+  ...output...</pre>
+</div>
+</div>
+<div class="section" id="determining-absolute-position-with-getposition">
+<h4><a class="toc-backref" href="#id21">Determining absolute position with getPosition()</a><a class="headerlink" href="#determining-absolute-position-with-getposition" title="Permalink to this headline">¶</a></h4>
+<p>Sometimes an option can affect or modify the meaning of another option. For
+example, consider <tt class="docutils literal"><span class="pre">gcc</span></tt>‘s <tt class="docutils literal"><span class="pre">-x</span> <span class="pre">LANG</span></tt> option. This tells <tt class="docutils literal"><span class="pre">gcc</span></tt> to ignore the
+suffix of subsequent positional arguments and force the file to be interpreted
+as if it contained source code in language <tt class="docutils literal"><span class="pre">LANG</span></tt>. In order to handle this
+properly, you need to know the absolute position of each argument, especially
+those in lists, so their interaction(s) can be applied correctly. This is also
+useful for options like <tt class="docutils literal"><span class="pre">-llibname</span></tt> which is actually a positional argument
+that starts with a dash.</p>
+<p>So, generally, the problem is that you have two <tt class="docutils literal"><span class="pre">cl::list</span></tt> variables that
+interact in some way. To ensure the correct interaction, you can use the
+<tt class="docutils literal"><span class="pre">cl::list::getPosition(optnum)</span></tt> method. This method returns the absolute
+position (as found on the command line) of the <tt class="docutils literal"><span class="pre">optnum</span></tt> item in the
+<tt class="docutils literal"><span class="pre">cl::list</span></tt>.</p>
+<p>The idiom for usage is like this:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">static</span> <span class="n">cl</span><span class="o">::</span><span class="n">list</span><span class="o"><</span><span class="n">std</span><span class="o">::</span><span class="n">string</span><span class="o">></span> <span class="n">Files</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Positional</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">OneOrMore</span><span class="p">);</span>
+<span class="k">static</span> <span class="n">cl</span><span class="o">::</span><span class="n">list</span><span class="o"><</span><span class="n">std</span><span class="o">::</span><span class="n">string</span><span class="o">></span> <span class="n">Libraries</span><span class="p">(</span><span class="s">"l"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">ZeroOrMore</span><span class="p">);</span>
+
+<span class="kt">int</span> <span class="nf">main</span><span class="p">(</span><span class="kt">int</span> <span class="n">argc</span><span class="p">,</span> <span class="kt">char</span><span class="o">**</span><span class="n">argv</span><span class="p">)</span> <span class="p">{</span>
+  <span class="c1">// ...</span>
+  <span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o"><</span><span class="n">std</span><span class="o">::</span><span class="n">string</span><span class="o">>::</span><span class="n">iterator</span> <span class="n">fileIt</span> <span class="o">=</span> <span class="n">Files</span><span class="p">.</span><span class="n">begin</span><span class="p">();</span>
+  <span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o"><</span><span class="n">std</span><span class="o">::</span><span class="n">string</span><span class="o">>::</span><span class="n">iterator</span> <span class="n">libIt</span>  <span class="o">=</span> <span class="n">Libraries</span><span class="p">.</span><span class="n">begin</span><span class="p">();</span>
+  <span class="kt">unsigned</span> <span class="n">libPos</span> <span class="o">=</span> <span class="mi">0</span><span class="p">,</span> <span class="n">filePos</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span>
+  <span class="k">while</span> <span class="p">(</span> <span class="mi">1</span> <span class="p">)</span> <span class="p">{</span>
+    <span class="k">if</span> <span class="p">(</span> <span class="n">libIt</span> <span class="o">!=</span> <span class="n">Libraries</span><span class="p">.</span><span class="n">end</span><span class="p">()</span> <span class="p">)</span>
+      <span class="n">libPos</span> <span class="o">=</span> <span class="n">Libraries</span><span class="p">.</span><span class="n">getPosition</span><span class="p">(</span> <span class="n">libIt</span> <span class="o">-</span> <span class="n">Libraries</span><span class="p">.</span><span class="n">begin</span><span class="p">()</span> <span class="p">);</span>
+    <span class="k">else</span>
+      <span class="n">libPos</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span>
+    <span class="k">if</span> <span class="p">(</span> <span class="n">fileIt</span> <span class="o">!=</span> <span class="n">Files</span><span class="p">.</span><span class="n">end</span><span class="p">()</span> <span class="p">)</span>
+      <span class="n">filePos</span> <span class="o">=</span> <span class="n">Files</span><span class="p">.</span><span class="n">getPosition</span><span class="p">(</span> <span class="n">fileIt</span> <span class="o">-</span> <span class="n">Files</span><span class="p">.</span><span class="n">begin</span><span class="p">()</span> <span class="p">);</span>
+    <span class="k">else</span>
+      <span class="n">filePos</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span>
+
+    <span class="k">if</span> <span class="p">(</span> <span class="n">filePos</span> <span class="o">!=</span> <span class="mi">0</span> <span class="o">&&</span> <span class="p">(</span><span class="n">libPos</span> <span class="o">==</span> <span class="mi">0</span> <span class="o">||</span> <span class="n">filePos</span> <span class="o"><</span> <span class="n">libPos</span><span class="p">)</span> <span class="p">)</span> <span class="p">{</span>
+      <span class="c1">// Source File Is next</span>
+      <span class="o">++</span><span class="n">fileIt</span><span class="p">;</span>
+    <span class="p">}</span>
+    <span class="k">else</span> <span class="k">if</span> <span class="p">(</span> <span class="n">libPos</span> <span class="o">!=</span> <span class="mi">0</span> <span class="o">&&</span> <span class="p">(</span><span class="n">filePos</span> <span class="o">==</span> <span class="mi">0</span> <span class="o">||</span> <span class="n">libPos</span> <span class="o"><</span> <span class="n">filePos</span><span class="p">)</span> <span class="p">)</span> <span class="p">{</span>
+      <span class="c1">// Library is next</span>
+      <span class="o">++</span><span class="n">libIt</span><span class="p">;</span>
+    <span class="p">}</span>
+    <span class="k">else</span>
+      <span class="k">break</span><span class="p">;</span> <span class="c1">// we're done with the list</span>
+  <span class="p">}</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>Note that, for compatibility reasons, the <tt class="docutils literal"><span class="pre">cl::opt</span></tt> also supports an
+<tt class="docutils literal"><span class="pre">unsigned</span> <span class="pre">getPosition()</span></tt> option that will provide the absolute position of
+that option. You can apply the same approach as above with a <tt class="docutils literal"><span class="pre">cl::opt</span></tt> and a
+<tt class="docutils literal"><span class="pre">cl::list</span></tt> option as you can with two lists.</p>
+</div>
+<div class="section" id="the-cl-consumeafter-modifier">
+<span id="this-section-for-more-information"></span><span id="cl-consumeafter"></span><span id="interpreter-style-options"></span><h4><a class="toc-backref" href="#id22">The <tt class="docutils literal"><span class="pre">cl::ConsumeAfter</span></tt> modifier</a><a class="headerlink" href="#the-cl-consumeafter-modifier" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::ConsumeAfter</span></tt> <a class="reference internal" href="#formatting-option">formatting option</a> is used to construct programs that
+use “interpreter style” option processing.  With this style of option
+processing, all arguments specified after the last positional argument are
+treated as special interpreter arguments that are not interpreted by the command
+line argument.</p>
+<p>As a concrete example, lets say we are developing a replacement for the standard
+Unix Bourne shell (<tt class="docutils literal"><span class="pre">/bin/sh</span></tt>).  To run <tt class="docutils literal"><span class="pre">/bin/sh</span></tt>, first you specify options
+to the shell itself (like <tt class="docutils literal"><span class="pre">-x</span></tt> which turns on trace output), then you specify
+the name of the script to run, then you specify arguments to the script.  These
+arguments to the script are parsed by the Bourne shell command line option
+processor, but are not interpreted as options to the shell itself.  Using the
+CommandLine library, we would specify this as:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="n">string</span><span class="o">></span> <span class="n">Script</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Positional</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"<input script>"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">init</span><span class="p">(</span><span class="s">"-"</span><span class="p">));</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">list</span><span class="o"><</span><span class="n">string</span><span class="o">></span>  <span class="n">Argv</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">ConsumeAfter</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"<program arguments>..."</span><span class="p">));</span>
+<span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span>    <span class="n">Trace</span><span class="p">(</span><span class="s">"x"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Enable trace output"</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>which automatically provides the help output:</p>
+<div class="highlight-python"><pre>USAGE: spiffysh [options] <input script> <program arguments>...
+
+OPTIONS:
+  -help - display available options (-help-hidden for more)
+  -x    - Enable trace output</pre>
+</div>
+<p>At runtime, if we run our new shell replacement as <tt class="docutils literal"><span class="pre">`spiffysh</span> <span class="pre">-x</span> <span class="pre">test.sh</span> <span class="pre">-a</span> <span class="pre">-x</span>
+<span class="pre">-y</span> <span class="pre">bar</span></tt>‘, the <tt class="docutils literal"><span class="pre">Trace</span></tt> variable will be set to true, the <tt class="docutils literal"><span class="pre">Script</span></tt> variable
+will be set to “<tt class="docutils literal"><span class="pre">test.sh</span></tt>”, and the <tt class="docutils literal"><span class="pre">Argv</span></tt> list will contain <tt class="docutils literal"><span class="pre">["-a",</span> <span class="pre">"-x",</span>
+<span class="pre">"-y",</span> <span class="pre">"bar"]</span></tt>, because they were specified after the last positional argument
+(which is the script name).</p>
+<p>There are several limitations to when <tt class="docutils literal"><span class="pre">cl::ConsumeAfter</span></tt> options can be
+specified.  For example, only one <tt class="docutils literal"><span class="pre">cl::ConsumeAfter</span></tt> can be specified per
+program, there must be at least one <a class="reference internal" href="#positional-argument">positional argument</a> specified, there must
+not be any <a class="reference internal" href="#cl-list">cl::list</a> positional arguments, and the <tt class="docutils literal"><span class="pre">cl::ConsumeAfter</span></tt> option
+should be a <a class="reference internal" href="#cl-list">cl::list</a> option.</p>
+</div>
+</div>
+<div class="section" id="internal-vs-external-storage">
+<span id="can-be-changed"></span><span id="id4"></span><h3><a class="toc-backref" href="#id23">Internal vs External Storage</a><a class="headerlink" href="#internal-vs-external-storage" title="Permalink to this headline">¶</a></h3>
+<p>By default, all command line options automatically hold the value that they
+parse from the command line.  This is very convenient in the common case,
+especially when combined with the ability to define command line options in the
+files that use them.  This is called the internal storage model.</p>
+<p>Sometimes, however, it is nice to separate the command line option processing
+code from the storage of the value parsed.  For example, lets say that we have a
+‘<tt class="docutils literal"><span class="pre">-debug</span></tt>‘ option that we would like to use to enable debug information across
+the entire body of our program.  In this case, the boolean value controlling the
+debug code should be globally accessible (in a header file, for example) yet the
+command line option processing code should not be exposed to all of these
+clients (requiring lots of .cpp files to <tt class="docutils literal"><span class="pre">#include</span> <span class="pre">CommandLine.h</span></tt>).</p>
+<p>To do this, set up your .h file with your option, like this for example:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="c1">// DebugFlag.h - Get access to the '-debug' command line option</span>
+<span class="c1">//</span>
+
+<span class="c1">// DebugFlag - This boolean is set to true if the '-debug' command line option</span>
+<span class="c1">// is specified.  This should probably not be referenced directly, instead, use</span>
+<span class="c1">// the DEBUG macro below.</span>
+<span class="c1">//</span>
+<span class="k">extern</span> <span class="kt">bool</span> <span class="n">DebugFlag</span><span class="p">;</span>
+
+<span class="c1">// DEBUG macro - This macro should be used by code to emit debug information.</span>
+<span class="c1">// In the '-debug' option is specified on the command line, and if this is a</span>
+<span class="c1">// debug build, then the code specified as the option to the macro will be</span>
+<span class="c1">// executed.  Otherwise it will not be.</span>
+<span class="cp">#ifdef NDEBUG</span>
+<span class="cp">#define DEBUG(X)</span>
+<span class="cp">#else</span>
+<span class="cp">#define DEBUG(X) do { if (DebugFlag) { X; } } while (0)</span>
+<span class="cp">#endif</span>
+</pre></div>
+</div>
+<p>This allows clients to blissfully use the <tt class="docutils literal"><span class="pre">DEBUG()</span></tt> macro, or the
+<tt class="docutils literal"><span class="pre">DebugFlag</span></tt> explicitly if they want to.  Now we just need to be able to set
+the <tt class="docutils literal"><span class="pre">DebugFlag</span></tt> boolean when the option is set.  To do this, we pass an
+additional argument to our command line argument processor, and we specify where
+to fill in with the <a class="reference internal" href="#cl-location">cl::location</a> attribute:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="kt">bool</span> <span class="n">DebugFlag</span><span class="p">;</span>                  <span class="c1">// the actual value</span>
+<span class="k">static</span> <span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="p">,</span> <span class="nb">true</span><span class="o">></span>       <span class="c1">// The parser</span>
+<span class="n">Debug</span><span class="p">(</span><span class="s">"debug"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Enable debug output"</span><span class="p">),</span> <span class="n">cl</span><span class="o">::</span><span class="n">Hidden</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">location</span><span class="p">(</span><span class="n">DebugFlag</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>In the above example, we specify “<tt class="docutils literal"><span class="pre">true</span></tt>” as the second argument to the
+<a class="reference internal" href="#cl-opt">cl::opt</a> template, indicating that the template should not maintain a copy of
+the value itself.  In addition to this, we specify the <a class="reference internal" href="#cl-location">cl::location</a>
+attribute, so that <tt class="docutils literal"><span class="pre">DebugFlag</span></tt> is automatically set.</p>
+</div>
+<div class="section" id="option-attributes">
+<h3><a class="toc-backref" href="#id24">Option Attributes</a><a class="headerlink" href="#option-attributes" title="Permalink to this headline">¶</a></h3>
+<p>This section describes the basic attributes that you can specify on options.</p>
+<ul>
+<li><p class="first">The option name attribute (which is required for all options, except
+<a class="reference internal" href="#positional-options">positional options</a>) specifies what the option name is.  This option is
+specified in simple double quotes:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">bool</span><span class="o">></span> <span class="n">Quiet</span><span class="p">(</span><span class="s">"quiet"</span><span class="p">);</span>
+</pre></div>
+</div>
+</li>
+</ul>
+<ul class="simple" id="cl-desc">
+<li>The <strong>cl::desc</strong> attribute specifies a description for the option to be
+shown in the <tt class="docutils literal"><span class="pre">-help</span></tt> output for the program. This attribute supports
+multi-line descriptions with lines separated by ‘n’.</li>
+</ul>
+<ul class="simple" id="cl-value-desc">
+<li>The <strong>cl::value_desc</strong> attribute specifies a string that can be used to
+fine tune the <tt class="docutils literal"><span class="pre">-help</span></tt> output for a command line option.  Look <a class="reference internal" href="#here">here</a> for an
+example.</li>
+</ul>
+<ul id="cl-init">
+<li><p class="first">The <strong>cl::init</strong> attribute specifies an initial value for a <a class="reference internal" href="#scalar">scalar</a>
+option.  If this attribute is not specified then the command line option value
+defaults to the value created by the default constructor for the
+type.</p>
+<div class="admonition warning">
+<p class="first admonition-title">Warning</p>
+<p class="last">If you specify both <strong>cl::init</strong> and <strong>cl::location</strong> for an option, you
+must specify <strong>cl::location</strong> first, so that when the command-line parser
+sees <strong>cl::init</strong>, it knows where to put the initial value. (You will get an
+error at runtime if you don’t put them in the right order.)</p>
+</div>
+</li>
+</ul>
+<ul class="simple" id="cl-location">
+<li>The <strong>cl::location</strong> attribute where to store the value for a parsed command
+line option if using external storage.  See the section on <a class="reference internal" href="#internal-vs-external-storage">Internal vs
+External Storage</a> for more information.</li>
+</ul>
+<ul class="simple" id="cl-aliasopt">
+<li>The <strong>cl::aliasopt</strong> attribute specifies which option a <a class="reference internal" href="#cl-alias">cl::alias</a> option is
+an alias for.</li>
+</ul>
+<ul id="cl-values">
+<li><p class="first">The <strong>cl::values</strong> attribute specifies the string-to-value mapping to be used
+by the generic parser.  It takes a <strong>clEnumValEnd terminated</strong> list of
+(option, value, description) triplets that specify the option name, the value
+mapped to, and the description shown in the <tt class="docutils literal"><span class="pre">-help</span></tt> for the tool.  Because
+the generic parser is used most frequently with enum values, two macros are
+often useful:</p>
+<ol class="arabic simple">
+<li>The <strong>clEnumVal</strong> macro is used as a nice simple way to specify a triplet
+for an enum.  This macro automatically makes the option name be the same as
+the enum name.  The first option to the macro is the enum, the second is
+the description for the command line option.</li>
+<li>The <strong>clEnumValN</strong> macro is used to specify macro options where the option
+name doesn’t equal the enum name.  For this macro, the first argument is
+the enum value, the second is the flag name, and the second is the
+description.</li>
+</ol>
+<p>You will get a compile time error if you try to use cl::values with a parser
+that does not support it.</p>
+</li>
+</ul>
+<ul class="simple" id="cl-multi-val">
+<li>The <strong>cl::multi_val</strong> attribute specifies that this option takes has multiple
+values (example: <tt class="docutils literal"><span class="pre">-sectalign</span> <span class="pre">segname</span> <span class="pre">sectname</span> <span class="pre">sectvalue</span></tt>). This attribute
+takes one unsigned argument - the number of values for the option. This
+attribute is valid only on <tt class="docutils literal"><span class="pre">cl::list</span></tt> options (and will fail with compile
+error if you try to use it with other option types). It is allowed to use all
+of the usual modifiers on multi-valued options (besides
+<tt class="docutils literal"><span class="pre">cl::ValueDisallowed</span></tt>, obviously).</li>
+</ul>
+<ul class="simple" id="cl-cat">
+<li>The <strong>cl::cat</strong> attribute specifies the option category that the option
+belongs to. The category should be a <a class="reference internal" href="#cl-optioncategory">cl::OptionCategory</a> object.</li>
+</ul>
+</div>
+<div class="section" id="option-modifiers">
+<h3><a class="toc-backref" href="#id25">Option Modifiers</a><a class="headerlink" href="#option-modifiers" title="Permalink to this headline">¶</a></h3>
+<p>Option modifiers are the flags and expressions that you pass into the
+constructors for <a class="reference internal" href="#cl-opt">cl::opt</a> and <a class="reference internal" href="#cl-list">cl::list</a>.  These modifiers give you the
+ability to tweak how options are parsed and how <tt class="docutils literal"><span class="pre">-help</span></tt> output is generated to
+fit your application well.</p>
+<p>These options fall into five main categories:</p>
+<ol class="arabic simple">
+<li>Hiding an option from <tt class="docutils literal"><span class="pre">-help</span></tt> output</li>
+<li>Controlling the number of occurrences required and allowed</li>
+<li>Controlling whether or not a value must be specified</li>
+<li>Controlling other formatting options</li>
+<li>Miscellaneous option modifiers</li>
+</ol>
+<p>It is not possible to specify two options from the same category (you’ll get a
+runtime error) to a single option, except for options in the miscellaneous
+category.  The CommandLine library specifies defaults for all of these settings
+that are the most useful in practice and the most common, which mean that you
+usually shouldn’t have to worry about these.</p>
+<div class="section" id="hiding-an-option-from-help-output">
+<h4><a class="toc-backref" href="#id26">Hiding an option from <tt class="docutils literal"><span class="pre">-help</span></tt> output</a><a class="headerlink" href="#hiding-an-option-from-help-output" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::NotHidden</span></tt>, <tt class="docutils literal"><span class="pre">cl::Hidden</span></tt>, and <tt class="docutils literal"><span class="pre">cl::ReallyHidden</span></tt> modifiers are
+used to control whether or not an option appears in the <tt class="docutils literal"><span class="pre">-help</span></tt> and
+<tt class="docutils literal"><span class="pre">-help-hidden</span></tt> output for the compiled program:</p>
+<ul class="simple" id="cl-nothidden">
+<li>The <strong>cl::NotHidden</strong> modifier (which is the default for <a class="reference internal" href="#cl-opt">cl::opt</a> and
+<a class="reference internal" href="#cl-list">cl::list</a> options) indicates the option is to appear in both help
+listings.</li>
+</ul>
+<ul class="simple" id="cl-hidden">
+<li>The <strong>cl::Hidden</strong> modifier (which is the default for <a class="reference internal" href="#cl-alias">cl::alias</a> options)
+indicates that the option should not appear in the <tt class="docutils literal"><span class="pre">-help</span></tt> output, but
+should appear in the <tt class="docutils literal"><span class="pre">-help-hidden</span></tt> output.</li>
+</ul>
+<ul class="simple" id="cl-reallyhidden">
+<li>The <strong>cl::ReallyHidden</strong> modifier indicates that the option should not appear
+in any help output.</li>
+</ul>
+</div>
+<div class="section" id="controlling-the-number-of-occurrences-required-and-allowed">
+<h4><a class="toc-backref" href="#id27">Controlling the number of occurrences required and allowed</a><a class="headerlink" href="#controlling-the-number-of-occurrences-required-and-allowed" title="Permalink to this headline">¶</a></h4>
+<p>This group of options is used to control how many time an option is allowed (or
+required) to be specified on the command line of your program.  Specifying a
+value for this setting allows the CommandLine library to do error checking for
+you.</p>
+<p>The allowed values for this option group are:</p>
+<ul class="simple" id="cl-optional">
+<li>The <strong>cl::Optional</strong> modifier (which is the default for the <a class="reference internal" href="#cl-opt">cl::opt</a> and
+<a class="reference internal" href="#cl-alias">cl::alias</a> classes) indicates that your program will allow either zero or
+one occurrence of the option to be specified.</li>
+</ul>
+<ul class="simple" id="cl-zeroormore">
+<li>The <strong>cl::ZeroOrMore</strong> modifier (which is the default for the <a class="reference internal" href="#cl-list">cl::list</a>
+class) indicates that your program will allow the option to be specified zero
+or more times.</li>
+</ul>
+<ul class="simple" id="cl-required">
+<li>The <strong>cl::Required</strong> modifier indicates that the specified option must be
+specified exactly one time.</li>
+</ul>
+<ul class="simple" id="cl-oneormore">
+<li>The <strong>cl::OneOrMore</strong> modifier indicates that the option must be specified at
+least one time.</li>
+<li>The <strong>cl::ConsumeAfter</strong> modifier is described in the <a class="reference internal" href="#positional-arguments-section">Positional arguments
+section</a>.</li>
+</ul>
+<p>If an option is not specified, then the value of the option is equal to the
+value specified by the <a class="reference internal" href="#cl-init">cl::init</a> attribute.  If the <tt class="docutils literal"><span class="pre">cl::init</span></tt> attribute is
+not specified, the option value is initialized with the default constructor for
+the data type.</p>
+<p>If an option is specified multiple times for an option of the <a class="reference internal" href="#cl-opt">cl::opt</a> class,
+only the last value will be retained.</p>
+</div>
+<div class="section" id="controlling-whether-or-not-a-value-must-be-specified">
+<h4><a class="toc-backref" href="#id28">Controlling whether or not a value must be specified</a><a class="headerlink" href="#controlling-whether-or-not-a-value-must-be-specified" title="Permalink to this headline">¶</a></h4>
+<p>This group of options is used to control whether or not the option allows a
+value to be present.  In the case of the CommandLine library, a value is either
+specified with an equal sign (e.g. ‘<tt class="docutils literal"><span class="pre">-index-depth=17</span></tt>‘) or as a trailing
+string (e.g. ‘<tt class="docutils literal"><span class="pre">-o</span> <span class="pre">a.out</span></tt>‘).</p>
+<p>The allowed values for this option group are:</p>
+<ul class="simple" id="cl-valueoptional">
+<li>The <strong>cl::ValueOptional</strong> modifier (which is the default for <tt class="docutils literal"><span class="pre">bool</span></tt> typed
+options) specifies that it is acceptable to have a value, or not.  A boolean
+argument can be enabled just by appearing on the command line, or it can have
+an explicit ‘<tt class="docutils literal"><span class="pre">-foo=true</span></tt>‘.  If an option is specified with this mode, it is
+illegal for the value to be provided without the equal sign.  Therefore
+‘<tt class="docutils literal"><span class="pre">-foo</span> <span class="pre">true</span></tt>‘ is illegal.  To get this behavior, you must use
+the <a class="reference internal" href="#cl-valuerequired">cl::ValueRequired</a> modifier.</li>
+</ul>
+<ul class="simple" id="cl-valuerequired">
+<li>The <strong>cl::ValueRequired</strong> modifier (which is the default for all other types
+except for <a class="reference internal" href="#unnamed-alternatives-using-the-generic-parser">unnamed alternatives using the generic parser</a>) specifies that a
+value must be provided.  This mode informs the command line library that if an
+option is not provides with an equal sign, that the next argument provided
+must be the value.  This allows things like ‘<tt class="docutils literal"><span class="pre">-o</span> <span class="pre">a.out</span></tt>‘ to work.</li>
+</ul>
+<ul class="simple" id="cl-valuedisallowed">
+<li>The <strong>cl::ValueDisallowed</strong> modifier (which is the default for <a class="reference internal" href="#unnamed-alternatives-using-the-generic-parser">unnamed
+alternatives using the generic parser</a>) indicates that it is a runtime error
+for the user to specify a value.  This can be provided to disallow users from
+providing options to boolean options (like ‘<tt class="docutils literal"><span class="pre">-foo=true</span></tt>‘).</li>
+</ul>
+<p>In general, the default values for this option group work just like you would
+want them to.  As mentioned above, you can specify the <a class="reference internal" href="#cl-valuedisallowed">cl::ValueDisallowed</a>
+modifier to a boolean argument to restrict your command line parser.  These
+options are mostly useful when <a class="reference internal" href="#extending-the-library">extending the library</a>.</p>
+</div>
+<div class="section" id="controlling-other-formatting-options">
+<span id="formatting-option"></span><h4><a class="toc-backref" href="#id29">Controlling other formatting options</a><a class="headerlink" href="#controlling-other-formatting-options" title="Permalink to this headline">¶</a></h4>
+<p>The formatting option group is used to specify that the command line option has
+special abilities and is otherwise different from other command line arguments.
+As usual, you can only specify one of these arguments at most.</p>
+<ul class="simple" id="cl-normalformatting">
+<li>The <strong>cl::NormalFormatting</strong> modifier (which is the default all options)
+specifies that this option is “normal”.</li>
+</ul>
+<ul class="simple" id="cl-positional">
+<li>The <strong>cl::Positional</strong> modifier specifies that this is a positional argument
+that does not have a command line option associated with it.  See the
+<a class="reference internal" href="#positional-arguments">Positional Arguments</a> section for more information.</li>
+<li>The <strong>cl::ConsumeAfter</strong> modifier specifies that this option is used to
+capture “interpreter style” arguments.  See <a class="reference internal" href="#this-section-for-more-information">this section for more
+information</a>.</li>
+</ul>
+<ul class="simple" id="cl-prefix">
+<span id="prefix"></span><li>The <strong>cl::Prefix</strong> modifier specifies that this option prefixes its value.
+With ‘Prefix’ options, the equal sign does not separate the value from the
+option name specified. Instead, the value is everything after the prefix,
+including any equal sign if present. This is useful for processing odd
+arguments like <tt class="docutils literal"><span class="pre">-lmalloc</span></tt> and <tt class="docutils literal"><span class="pre">-L/usr/lib</span></tt> in a linker tool or
+<tt class="docutils literal"><span class="pre">-DNAME=value</span></tt> in a compiler tool.  Here, the ‘<tt class="docutils literal"><span class="pre">l</span></tt>‘, ‘<tt class="docutils literal"><span class="pre">D</span></tt>‘ and ‘<tt class="docutils literal"><span class="pre">L</span></tt>‘
+options are normal string (or list) options, that have the <strong>cl::Prefix</strong>
+modifier added to allow the CommandLine library to recognize them.  Note that
+<strong>cl::Prefix</strong> options must not have the <strong>cl::ValueDisallowed</strong> modifier
+specified.</li>
+</ul>
+<ul class="simple" id="cl-grouping">
+<span id="grouping"></span><li>The <strong>cl::Grouping</strong> modifier is used to implement Unix-style tools (like
+<tt class="docutils literal"><span class="pre">ls</span></tt>) that have lots of single letter arguments, but only require a single
+dash.  For example, the ‘<tt class="docutils literal"><span class="pre">ls</span> <span class="pre">-labF</span></tt>‘ command actually enables four different
+options, all of which are single letters.  Note that <strong>cl::Grouping</strong> options
+cannot have values.</li>
+</ul>
+<p>The CommandLine library does not restrict how you use the <strong>cl::Prefix</strong> or
+<strong>cl::Grouping</strong> modifiers, but it is possible to specify ambiguous argument
+settings.  Thus, it is possible to have multiple letter options that are prefix
+or grouping options, and they will still work as designed.</p>
+<p>To do this, the CommandLine library uses a greedy algorithm to parse the input
+option into (potentially multiple) prefix and grouping options.  The strategy
+basically looks like this:</p>
+<div class="highlight-python"><pre>parse(string OrigInput) {
+
+1. string input = OrigInput;
+2. if (isOption(input)) return getOption(input).parse();  // Normal option
+3. while (!isOption(input) && !input.empty()) input.pop_back();  // Remove the last letter
+4. if (input.empty()) return error();  // No matching option
+5. if (getOption(input).isPrefix())
+     return getOption(input).parse(input);
+6. while (!input.empty()) {  // Must be grouping options
+     getOption(input).parse();
+     OrigInput.erase(OrigInput.begin(), OrigInput.begin()+input.length());
+     input = OrigInput;
+     while (!isOption(input) && !input.empty()) input.pop_back();
+   }
+7. if (!OrigInput.empty()) error();
+
+}</pre>
+</div>
+</div>
+<div class="section" id="miscellaneous-option-modifiers">
+<h4><a class="toc-backref" href="#id30">Miscellaneous option modifiers</a><a class="headerlink" href="#miscellaneous-option-modifiers" title="Permalink to this headline">¶</a></h4>
+<p>The miscellaneous option modifiers are the only flags where you can specify more
+than one flag from the set: they are not mutually exclusive.  These flags
+specify boolean properties that modify the option.</p>
+<ul class="simple" id="cl-commaseparated">
+<li>The <strong>cl::CommaSeparated</strong> modifier indicates that any commas specified for an
+option’s value should be used to split the value up into multiple values for
+the option.  For example, these two options are equivalent when
+<tt class="docutils literal"><span class="pre">cl::CommaSeparated</span></tt> is specified: “<tt class="docutils literal"><span class="pre">-foo=a</span> <span class="pre">-foo=b</span> <span class="pre">-foo=c</span></tt>” and
+“<tt class="docutils literal"><span class="pre">-foo=a,b,c</span></tt>”.  This option only makes sense to be used in a case where the
+option is allowed to accept one or more values (i.e. it is a <a class="reference internal" href="#cl-list">cl::list</a>
+option).</li>
+</ul>
+<ul class="simple" id="cl-positionaleatsargs">
+<li>The <strong>cl::PositionalEatsArgs</strong> modifier (which only applies to positional
+arguments, and only makes sense for lists) indicates that positional argument
+should consume any strings after it (including strings that start with a “-”)
+up until another recognized positional argument.  For example, if you have two
+“eating” positional arguments, “<tt class="docutils literal"><span class="pre">pos1</span></tt>” and “<tt class="docutils literal"><span class="pre">pos2</span></tt>”, the string “<tt class="docutils literal"><span class="pre">-pos1</span>
+<span class="pre">-foo</span> <span class="pre">-bar</span> <span class="pre">baz</span> <span class="pre">-pos2</span> <span class="pre">-bork</span></tt>” would cause the “<tt class="docutils literal"><span class="pre">-foo</span> <span class="pre">-bar</span> <span class="pre">-baz</span></tt>” strings to
+be applied to the “<tt class="docutils literal"><span class="pre">-pos1</span></tt>” option and the “<tt class="docutils literal"><span class="pre">-bork</span></tt>” string to be applied
+to the “<tt class="docutils literal"><span class="pre">-pos2</span></tt>” option.</li>
+</ul>
+<ul class="simple" id="cl-sink">
+<li>The <strong>cl::Sink</strong> modifier is used to handle unknown options. If there is at
+least one option with <tt class="docutils literal"><span class="pre">cl::Sink</span></tt> modifier specified, the parser passes
+unrecognized option strings to it as values instead of signaling an error. As
+with <tt class="docutils literal"><span class="pre">cl::CommaSeparated</span></tt>, this modifier only makes sense with a <a class="reference internal" href="#cl-list">cl::list</a>
+option.</li>
+</ul>
+<p>So far, these are the only three miscellaneous option modifiers.</p>
+</div>
+<div class="section" id="response-files">
+<span id="id5"></span><h4><a class="toc-backref" href="#id31">Response files</a><a class="headerlink" href="#response-files" title="Permalink to this headline">¶</a></h4>
+<p>Some systems, such as certain variants of Microsoft Windows and some older
+Unices have a relatively low limit on command-line length. It is therefore
+customary to use the so-called ‘response files’ to circumvent this
+restriction. These files are mentioned on the command-line (using the “@file”)
+syntax. The program reads these files and inserts the contents into argv,
+thereby working around the command-line length limits. Response files are
+enabled by an optional fourth argument to <a class="reference internal" href="#cl-parseenvironmentoptions">cl::ParseEnvironmentOptions</a> and
+<a class="reference internal" href="#cl-parsecommandlineoptions">cl::ParseCommandLineOptions</a>.</p>
+</div>
+</div>
+<div class="section" id="top-level-classes-and-functions">
+<h3><a class="toc-backref" href="#id32">Top-Level Classes and Functions</a><a class="headerlink" href="#top-level-classes-and-functions" title="Permalink to this headline">¶</a></h3>
+<p>Despite all of the built-in flexibility, the CommandLine option library really
+only consists of one function <a class="reference internal" href="#cl-parsecommandlineoptions">cl::ParseCommandLineOptions</a>) and three main
+classes: <a class="reference internal" href="#cl-opt">cl::opt</a>, <a class="reference internal" href="#cl-list">cl::list</a>, and <a class="reference internal" href="#cl-alias">cl::alias</a>.  This section describes
+these three classes in detail.</p>
+<div class="section" id="the-cl-getregisteredoptions-function">
+<span id="cl-getregisteredoptions"></span><h4><a class="toc-backref" href="#id33">The <tt class="docutils literal"><span class="pre">cl::getRegisteredOptions</span></tt> function</a><a class="headerlink" href="#the-cl-getregisteredoptions-function" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::getRegisteredOptions</span></tt> function is designed to give a programmer
+access to declared non-positional command line options so that how they appear
+in <tt class="docutils literal"><span class="pre">-help</span></tt> can be modified prior to calling <a class="reference internal" href="#cl-parsecommandlineoptions">cl::ParseCommandLineOptions</a>.
+Note this method should not be called during any static initialisation because
+it cannot be guaranteed that all options will have been initialised. Hence it
+should be called from <tt class="docutils literal"><span class="pre">main</span></tt>.</p>
+<p>This function can be used to gain access to options declared in libraries that
+the tool writter may not have direct access to.</p>
+<p>The function retrieves a <a class="reference internal" href="ProgrammersManual.html#dss-stringmap"><em>StringMap</em></a> that maps the option
+string (e.g. <tt class="docutils literal"><span class="pre">-help</span></tt>) to an <tt class="docutils literal"><span class="pre">Option*</span></tt>.</p>
+<p>Here is an example of how the function could be used:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">using</span> <span class="k">namespace</span> <span class="n">llvm</span><span class="p">;</span>
+<span class="kt">int</span> <span class="nf">main</span><span class="p">(</span><span class="kt">int</span> <span class="n">argc</span><span class="p">,</span> <span class="kt">char</span> <span class="o">**</span><span class="n">argv</span><span class="p">)</span> <span class="p">{</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">OptionCategory</span> <span class="n">AnotherCategory</span><span class="p">(</span><span class="s">"Some options"</span><span class="p">);</span>
+
+  <span class="n">StringMap</span><span class="o"><</span><span class="n">cl</span><span class="o">::</span><span class="n">Option</span><span class="o">*></span> <span class="n">Map</span><span class="p">;</span>
+  <span class="n">cl</span><span class="o">::</span><span class="n">getRegisteredOptions</span><span class="p">(</span><span class="n">Map</span><span class="p">);</span>
+
+  <span class="c1">//Unhide useful option and put it in a different category</span>
+  <span class="n">assert</span><span class="p">(</span><span class="n">Map</span><span class="p">.</span><span class="n">count</span><span class="p">(</span><span class="s">"print-all-options"</span><span class="p">)</span> <span class="o">></span> <span class="mi">0</span><span class="p">);</span>
+  <span class="n">Map</span><span class="p">[</span><span class="s">"print-all-options"</span><span class="p">]</span><span class="o">-></span><span class="n">setHiddenFlag</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">NotHidden</span><span class="p">);</span>
+  <span class="n">Map</span><span class="p">[</span><span class="s">"print-all-options"</span><span class="p">]</span><span class="o">-></span><span class="n">setCategory</span><span class="p">(</span><span class="n">AnotherCategory</span><span class="p">);</span>
+
+  <span class="c1">//Hide an option we don't want to see</span>
+  <span class="n">assert</span><span class="p">(</span><span class="n">Map</span><span class="p">.</span><span class="n">count</span><span class="p">(</span><span class="s">"enable-no-infs-fp-math"</span><span class="p">)</span> <span class="o">></span> <span class="mi">0</span><span class="p">);</span>
+  <span class="n">Map</span><span class="p">[</span><span class="s">"enable-no-infs-fp-math"</span><span class="p">]</span><span class="o">-></span><span class="n">setHiddenFlag</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Hidden</span><span class="p">);</span>
+
+  <span class="c1">//Change --version to --show-version</span>
+  <span class="n">assert</span><span class="p">(</span><span class="n">Map</span><span class="p">.</span><span class="n">count</span><span class="p">(</span><span class="s">"version"</span><span class="p">)</span> <span class="o">></span> <span class="mi">0</span><span class="p">);</span>
+  <span class="n">Map</span><span class="p">[</span><span class="s">"version"</span><span class="p">]</span><span class="o">-></span><span class="n">setArgStr</span><span class="p">(</span><span class="s">"show-version"</span><span class="p">);</span>
+
+  <span class="c1">//Change --help description</span>
+  <span class="n">assert</span><span class="p">(</span><span class="n">Map</span><span class="p">.</span><span class="n">count</span><span class="p">(</span><span class="s">"help"</span><span class="p">)</span> <span class="o">></span> <span class="mi">0</span><span class="p">);</span>
+  <span class="n">Map</span><span class="p">[</span><span class="s">"help"</span><span class="p">]</span><span class="o">-></span><span class="n">setDescription</span><span class="p">(</span><span class="s">"Shows help"</span><span class="p">);</span>
+
+  <span class="n">cl</span><span class="o">::</span><span class="n">ParseCommandLineOptions</span><span class="p">(</span><span class="n">argc</span><span class="p">,</span> <span class="n">argv</span><span class="p">,</span> <span class="s">"This is a small program to demo the LLVM CommandLine API"</span><span class="p">);</span>
+  <span class="p">...</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="the-cl-parsecommandlineoptions-function">
+<span id="cl-parsecommandlineoptions"></span><h4><a class="toc-backref" href="#id34">The <tt class="docutils literal"><span class="pre">cl::ParseCommandLineOptions</span></tt> function</a><a class="headerlink" href="#the-cl-parsecommandlineoptions-function" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::ParseCommandLineOptions</span></tt> function is designed to be called directly
+from <tt class="docutils literal"><span class="pre">main</span></tt>, and is used to fill in the values of all of the command line
+option variables once <tt class="docutils literal"><span class="pre">argc</span></tt> and <tt class="docutils literal"><span class="pre">argv</span></tt> are available.</p>
+<p>The <tt class="docutils literal"><span class="pre">cl::ParseCommandLineOptions</span></tt> function requires two parameters (<tt class="docutils literal"><span class="pre">argc</span></tt>
+and <tt class="docutils literal"><span class="pre">argv</span></tt>), but may also take an optional third parameter which holds
+<a class="reference internal" href="#additional-extra-text">additional extra text</a> to emit when the <tt class="docutils literal"><span class="pre">-help</span></tt> option is invoked, and a
+fourth boolean parameter that enables <a class="reference internal" href="#response-files">response files</a>.</p>
+</div>
+<div class="section" id="the-cl-parseenvironmentoptions-function">
+<span id="cl-parseenvironmentoptions"></span><h4><a class="toc-backref" href="#id35">The <tt class="docutils literal"><span class="pre">cl::ParseEnvironmentOptions</span></tt> function</a><a class="headerlink" href="#the-cl-parseenvironmentoptions-function" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::ParseEnvironmentOptions</span></tt> function has mostly the same effects as
+<a class="reference internal" href="#cl-parsecommandlineoptions">cl::ParseCommandLineOptions</a>, except that it is designed to take values for
+options from an environment variable, for those cases in which reading the
+command line is not convenient or desired. It fills in the values of all the
+command line option variables just like <a class="reference internal" href="#cl-parsecommandlineoptions">cl::ParseCommandLineOptions</a> does.</p>
+<p>It takes four parameters: the name of the program (since <tt class="docutils literal"><span class="pre">argv</span></tt> may not be
+available, it can’t just look in <tt class="docutils literal"><span class="pre">argv[0]</span></tt>), the name of the environment
+variable to examine, the optional <a class="reference internal" href="#additional-extra-text">additional extra text</a> to emit when the
+<tt class="docutils literal"><span class="pre">-help</span></tt> option is invoked, and the boolean switch that controls whether
+<a class="reference internal" href="#response-files">response files</a> should be read.</p>
+<p><tt class="docutils literal"><span class="pre">cl::ParseEnvironmentOptions</span></tt> will break the environment variable’s value up
+into words and then process them using <a class="reference internal" href="#cl-parsecommandlineoptions">cl::ParseCommandLineOptions</a>.
+<strong>Note:</strong> Currently <tt class="docutils literal"><span class="pre">cl::ParseEnvironmentOptions</span></tt> does not support quoting, so
+an environment variable containing <tt class="docutils literal"><span class="pre">-option</span> <span class="pre">"foo</span> <span class="pre">bar"</span></tt> will be parsed as three
+words, <tt class="docutils literal"><span class="pre">-option</span></tt>, <tt class="docutils literal"><span class="pre">"foo</span></tt>, and <tt class="docutils literal"><span class="pre">bar"</span></tt>, which is different from what you
+would get from the shell with the same input.</p>
+</div>
+<div class="section" id="the-cl-setversionprinter-function">
+<h4><a class="toc-backref" href="#id36">The <tt class="docutils literal"><span class="pre">cl::SetVersionPrinter</span></tt> function</a><a class="headerlink" href="#the-cl-setversionprinter-function" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::SetVersionPrinter</span></tt> function is designed to be called directly from
+<tt class="docutils literal"><span class="pre">main</span></tt> and <em>before</em> <tt class="docutils literal"><span class="pre">cl::ParseCommandLineOptions</span></tt>. Its use is optional. It
+simply arranges for a function to be called in response to the <tt class="docutils literal"><span class="pre">--version</span></tt>
+option instead of having the <tt class="docutils literal"><span class="pre">CommandLine</span></tt> library print out the usual version
+string for LLVM. This is useful for programs that are not part of LLVM but wish
+to use the <tt class="docutils literal"><span class="pre">CommandLine</span></tt> facilities. Such programs should just define a small
+function that takes no arguments and returns <tt class="docutils literal"><span class="pre">void</span></tt> and that prints out
+whatever version information is appropriate for the program. Pass the address of
+that function to <tt class="docutils literal"><span class="pre">cl::SetVersionPrinter</span></tt> to arrange for it to be called when
+the <tt class="docutils literal"><span class="pre">--version</span></tt> option is given by the user.</p>
+</div>
+<div class="section" id="the-cl-opt-class">
+<span id="scalar"></span><span id="cl-opt"></span><h4><a class="toc-backref" href="#id37">The <tt class="docutils literal"><span class="pre">cl::opt</span></tt> class</a><a class="headerlink" href="#the-cl-opt-class" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::opt</span></tt> class is the class used to represent scalar command line
+options, and is the one used most of the time.  It is a templated class which
+can take up to three arguments (all except for the first have default values
+though):</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">namespace</span> <span class="n">cl</span> <span class="p">{</span>
+  <span class="k">template</span> <span class="o"><</span><span class="k">class</span> <span class="nc">DataType</span><span class="p">,</span> <span class="kt">bool</span> <span class="n">ExternalStorage</span> <span class="o">=</span> <span class="nb">false</span><span class="p">,</span>
+            <span class="k">class</span> <span class="nc">ParserClass</span> <span class="o">=</span> <span class="n">parser</span><span class="o"><</span><span class="n">DataType</span><span class="o">></span> <span class="o">></span>
+  <span class="k">class</span> <span class="nc">opt</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>The first template argument specifies what underlying data type the command line
+argument is, and is used to select a default parser implementation.  The second
+template argument is used to specify whether the option should contain the
+storage for the option (the default) or whether external storage should be used
+to contain the value parsed for the option (see <a class="reference internal" href="#internal-vs-external-storage">Internal vs External Storage</a>
+for more information).</p>
+<p>The third template argument specifies which parser to use.  The default value
+selects an instantiation of the <tt class="docutils literal"><span class="pre">parser</span></tt> class based on the underlying data
+type of the option.  In general, this default works well for most applications,
+so this option is only used when using a <a class="reference internal" href="#custom-parser">custom parser</a>.</p>
+</div>
+<div class="section" id="the-cl-list-class">
+<span id="cl-list"></span><span id="lists-of-arguments"></span><h4><a class="toc-backref" href="#id38">The <tt class="docutils literal"><span class="pre">cl::list</span></tt> class</a><a class="headerlink" href="#the-cl-list-class" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::list</span></tt> class is the class used to represent a list of command line
+options.  It too is a templated class which can take up to three arguments:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">namespace</span> <span class="n">cl</span> <span class="p">{</span>
+  <span class="k">template</span> <span class="o"><</span><span class="k">class</span> <span class="nc">DataType</span><span class="p">,</span> <span class="k">class</span> <span class="nc">Storage</span> <span class="o">=</span> <span class="kt">bool</span><span class="p">,</span>
+            <span class="k">class</span> <span class="nc">ParserClass</span> <span class="o">=</span> <span class="n">parser</span><span class="o"><</span><span class="n">DataType</span><span class="o">></span> <span class="o">></span>
+  <span class="k">class</span> <span class="nc">list</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>This class works the exact same as the <a class="reference internal" href="#cl-opt">cl::opt</a> class, except that the second
+argument is the <strong>type</strong> of the external storage, not a boolean value.  For this
+class, the marker type ‘<tt class="docutils literal"><span class="pre">bool</span></tt>‘ is used to indicate that internal storage
+should be used.</p>
+</div>
+<div class="section" id="the-cl-bits-class">
+<span id="cl-bits"></span><h4><a class="toc-backref" href="#id39">The <tt class="docutils literal"><span class="pre">cl::bits</span></tt> class</a><a class="headerlink" href="#the-cl-bits-class" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::bits</span></tt> class is the class used to represent a list of command line
+options in the form of a bit vector.  It is also a templated class which can
+take up to three arguments:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">namespace</span> <span class="n">cl</span> <span class="p">{</span>
+  <span class="k">template</span> <span class="o"><</span><span class="k">class</span> <span class="nc">DataType</span><span class="p">,</span> <span class="k">class</span> <span class="nc">Storage</span> <span class="o">=</span> <span class="kt">bool</span><span class="p">,</span>
+            <span class="k">class</span> <span class="nc">ParserClass</span> <span class="o">=</span> <span class="n">parser</span><span class="o"><</span><span class="n">DataType</span><span class="o">></span> <span class="o">></span>
+  <span class="k">class</span> <span class="nc">bits</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>This class works the exact same as the <a class="reference internal" href="#cl-list">cl::list</a> class, except that the second
+argument must be of <strong>type</strong> <tt class="docutils literal"><span class="pre">unsigned</span></tt> if external storage is used.</p>
+</div>
+<div class="section" id="the-cl-alias-class">
+<span id="cl-alias"></span><h4><a class="toc-backref" href="#id40">The <tt class="docutils literal"><span class="pre">cl::alias</span></tt> class</a><a class="headerlink" href="#the-cl-alias-class" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::alias</span></tt> class is a nontemplated class that is used to form aliases for
+other arguments.</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">namespace</span> <span class="n">cl</span> <span class="p">{</span>
+  <span class="k">class</span> <span class="nc">alias</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>The <a class="reference internal" href="#cl-aliasopt">cl::aliasopt</a> attribute should be used to specify which option this is an
+alias for.  Alias arguments default to being <a class="reference internal" href="#cl-hidden">cl::Hidden</a>, and use the aliased
+options parser to do the conversion from string to data.</p>
+</div>
+<div class="section" id="the-cl-extrahelp-class">
+<span id="cl-extrahelp"></span><h4><a class="toc-backref" href="#id41">The <tt class="docutils literal"><span class="pre">cl::extrahelp</span></tt> class</a><a class="headerlink" href="#the-cl-extrahelp-class" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::extrahelp</span></tt> class is a nontemplated class that allows extra help text
+to be printed out for the <tt class="docutils literal"><span class="pre">-help</span></tt> option.</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">namespace</span> <span class="n">cl</span> <span class="p">{</span>
+  <span class="k">struct</span> <span class="n">extrahelp</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>To use the extrahelp, simply construct one with a <tt class="docutils literal"><span class="pre">const</span> <span class="pre">char*</span></tt> parameter to
+the constructor. The text passed to the constructor will be printed at the
+bottom of the help message, verbatim. Note that multiple <tt class="docutils literal"><span class="pre">cl::extrahelp</span></tt>
+<strong>can</strong> be used, but this practice is discouraged. If your tool needs to print
+additional help information, put all that help into a single <tt class="docutils literal"><span class="pre">cl::extrahelp</span></tt>
+instance.</p>
+<p>For example:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">cl</span><span class="o">::</span><span class="n">extrahelp</span><span class="p">(</span><span class="s">"</span><span class="se">\n</span><span class="s">ADDITIONAL HELP:</span><span class="se">\n\n</span><span class="s">  This is the extra help</span><span class="se">\n</span><span class="s">"</span><span class="p">);</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="the-cl-optioncategory-class">
+<span id="cl-optioncategory"></span><h4><a class="toc-backref" href="#id42">The <tt class="docutils literal"><span class="pre">cl::OptionCategory</span></tt> class</a><a class="headerlink" href="#the-cl-optioncategory-class" title="Permalink to this headline">¶</a></h4>
+<p>The <tt class="docutils literal"><span class="pre">cl::OptionCategory</span></tt> class is a simple class for declaring
+option categories.</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">namespace</span> <span class="n">cl</span> <span class="p">{</span>
+  <span class="k">class</span> <span class="nc">OptionCategory</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>An option category must have a name and optionally a description which are
+passed to the constructor as <tt class="docutils literal"><span class="pre">const</span> <span class="pre">char*</span></tt>.</p>
+<p>Note that declaring an option category and associating it with an option before
+parsing options (e.g. statically) will change the output of <tt class="docutils literal"><span class="pre">-help</span></tt> from
+uncategorized to categorized. If an option category is declared but not
+associated with an option then it will be hidden from the output of <tt class="docutils literal"><span class="pre">-help</span></tt>
+but will be shown in the output of <tt class="docutils literal"><span class="pre">-help-hidden</span></tt>.</p>
+</div>
+</div>
+<div class="section" id="builtin-parsers">
+<span id="discussed-previously"></span><span id="different-parser"></span><h3><a class="toc-backref" href="#id43">Builtin parsers</a><a class="headerlink" href="#builtin-parsers" title="Permalink to this headline">¶</a></h3>
+<p>Parsers control how the string value taken from the command line is translated
+into a typed value, suitable for use in a C++ program.  By default, the
+CommandLine library uses an instance of <tt class="docutils literal"><span class="pre">parser<type></span></tt> if the command line
+option specifies that it uses values of type ‘<tt class="docutils literal"><span class="pre">type</span></tt>‘.  Because of this,
+custom option processing is specified with specializations of the ‘<tt class="docutils literal"><span class="pre">parser</span></tt>‘
+class.</p>
+<p>The CommandLine library provides the following builtin parser specializations,
+which are sufficient for most applications. It can, however, also be extended to
+work with new data types and new ways of interpreting the same data.  See the
+<a class="reference internal" href="#writing-a-custom-parser">Writing a Custom Parser</a> for more details on this type of library extension.</p>
+<ul class="simple" id="cl-parser">
+<span id="enums"></span><li>The generic <tt class="docutils literal"><span class="pre">parser<t></span></tt> parser can be used to map strings values to any data
+type, through the use of the <a class="reference internal" href="#cl-values">cl::values</a> property, which specifies the
+mapping information.  The most common use of this parser is for parsing enum
+values, which allows you to use the CommandLine library for all of the error
+checking to make sure that only valid enum values are specified (as opposed to
+accepting arbitrary strings).  Despite this, however, the generic parser class
+can be used for any data type.</li>
+</ul>
+<ul class="simple" id="bool-parser">
+<span id="boolean-flags"></span><li>The <strong>parser<bool> specialization</strong> is used to convert boolean strings to a
+boolean value.  Currently accepted strings are “<tt class="docutils literal"><span class="pre">true</span></tt>”, “<tt class="docutils literal"><span class="pre">TRUE</span></tt>”,
+“<tt class="docutils literal"><span class="pre">True</span></tt>”, “<tt class="docutils literal"><span class="pre">1</span></tt>”, “<tt class="docutils literal"><span class="pre">false</span></tt>”, “<tt class="docutils literal"><span class="pre">FALSE</span></tt>”, “<tt class="docutils literal"><span class="pre">False</span></tt>”, and “<tt class="docutils literal"><span class="pre">0</span></tt>”.</li>
+<li>The <strong>parser<boolOrDefault> specialization</strong> is used for cases where the value
+is boolean, but we also need to know whether the option was specified at all.
+boolOrDefault is an enum with 3 values, BOU_UNSET, BOU_TRUE and BOU_FALSE.
+This parser accepts the same strings as <strong>``parser<bool>``</strong>.</li>
+</ul>
+<ul class="simple" id="strings">
+<li>The <strong>parser<string> specialization</strong> simply stores the parsed string into the
+string value specified.  No conversion or modification of the data is
+performed.</li>
+</ul>
+<ul class="simple" id="int">
+<span id="integers"></span><li>The <strong>parser<int> specialization</strong> uses the C <tt class="docutils literal"><span class="pre">strtol</span></tt> function to parse the
+string input.  As such, it will accept a decimal number (with an optional ‘+’
+or ‘-‘ prefix) which must start with a non-zero digit.  It accepts octal
+numbers, which are identified with a ‘<tt class="docutils literal"><span class="pre">0</span></tt>‘ prefix digit, and hexadecimal
+numbers with a prefix of ‘<tt class="docutils literal"><span class="pre">0x</span></tt>‘ or ‘<tt class="docutils literal"><span class="pre">0X</span></tt>‘.</li>
+</ul>
+<ul class="simple" id="double">
+<span id="float"></span><span id="doubles"></span><li>The <strong>parser<double></strong> and <strong>parser<float> specializations</strong> use the standard
+C <tt class="docutils literal"><span class="pre">strtod</span></tt> function to convert floating point strings into floating point
+values.  As such, a broad range of string formats is supported, including
+exponential notation (ex: <tt class="docutils literal"><span class="pre">1.7e15</span></tt>) and properly supports locales.</li>
+</ul>
+</div>
+</div>
+<div class="section" id="extending-the-library">
+<span id="extension-guide"></span><span id="id6"></span><h2><a class="toc-backref" href="#id44">Extension Guide</a><a class="headerlink" href="#extending-the-library" title="Permalink to this headline">¶</a></h2>
+<p>Although the CommandLine library has a lot of functionality built into it
+already (as discussed previously), one of its true strengths lie in its
+extensibility.  This section discusses how the CommandLine library works under
+the covers and illustrates how to do some simple, common, extensions.</p>
+<div class="section" id="writing-a-custom-parser">
+<span id="custom-parser"></span><span id="custom-parsers"></span><span id="id7"></span><h3><a class="toc-backref" href="#id45">Writing a custom parser</a><a class="headerlink" href="#writing-a-custom-parser" title="Permalink to this headline">¶</a></h3>
+<p>One of the simplest and most common extensions is the use of a custom parser.
+As <a class="reference internal" href="#discussed-previously">discussed previously</a>, parsers are the portion of the CommandLine library
+that turns string input from the user into a particular parsed data type,
+validating the input in the process.</p>
+<p>There are two ways to use a new parser:</p>
+<ol class="arabic">
+<li><p class="first">Specialize the <a class="reference internal" href="#cl-parser">cl::parser</a> template for your custom data type.</p>
+<p>This approach has the advantage that users of your custom data type will
+automatically use your custom parser whenever they define an option with a
+value type of your data type.  The disadvantage of this approach is that it
+doesn’t work if your fundamental data type is something that is already
+supported.</p>
+</li>
+<li><p class="first">Write an independent class, using it explicitly from options that need it.</p>
+<p>This approach works well in situations where you would line to parse an
+option using special syntax for a not-very-special data-type.  The drawback
+of this approach is that users of your parser have to be aware that they are
+using your parser instead of the builtin ones.</p>
+</li>
+</ol>
+<p>To guide the discussion, we will discuss a custom parser that accepts file
+sizes, specified with an optional unit after the numeric size.  For example, we
+would like to parse “102kb”, “41M”, “1G” into the appropriate integer value.  In
+this case, the underlying data type we want to parse into is ‘<tt class="docutils literal"><span class="pre">unsigned</span></tt>‘.  We
+choose approach #2 above because we don’t want to make this the default for all
+<tt class="docutils literal"><span class="pre">unsigned</span></tt> options.</p>
+<p>To start out, we declare our new <tt class="docutils literal"><span class="pre">FileSizeParser</span></tt> class:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">struct</span> <span class="n">FileSizeParser</span> <span class="o">:</span> <span class="k">public</span> <span class="n">cl</span><span class="o">::</span><span class="n">basic_parser</span><span class="o"><</span><span class="kt">unsigned</span><span class="o">></span> <span class="p">{</span>
+  <span class="c1">// parse - Return true on error.</span>
+  <span class="kt">bool</span> <span class="n">parse</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Option</span> <span class="o">&</span><span class="n">O</span><span class="p">,</span> <span class="k">const</span> <span class="kt">char</span> <span class="o">*</span><span class="n">ArgName</span><span class="p">,</span> <span class="k">const</span> <span class="n">std</span><span class="o">::</span><span class="n">string</span> <span class="o">&</span><span class="n">ArgValue</span><span class="p">,</span>
+             <span class="kt">unsigned</span> <span class="o">&</span><span class="n">Val</span><span class="p">);</span>
+<span class="p">};</span>
+</pre></div>
+</div>
+<p>Our new class inherits from the <tt class="docutils literal"><span class="pre">cl::basic_parser</span></tt> template class to fill in
+the default, boiler plate code for us.  We give it the data type that we parse
+into, the last argument to the <tt class="docutils literal"><span class="pre">parse</span></tt> method, so that clients of our custom
+parser know what object type to pass in to the parse method.  (Here we declare
+that we parse into ‘<tt class="docutils literal"><span class="pre">unsigned</span></tt>‘ variables.)</p>
+<p>For most purposes, the only method that must be implemented in a custom parser
+is the <tt class="docutils literal"><span class="pre">parse</span></tt> method.  The <tt class="docutils literal"><span class="pre">parse</span></tt> method is called whenever the option is
+invoked, passing in the option itself, the option name, the string to parse, and
+a reference to a return value.  If the string to parse is not well-formed, the
+parser should output an error message and return true.  Otherwise it should
+return false and set ‘<tt class="docutils literal"><span class="pre">Val</span></tt>‘ to the parsed value.  In our example, we
+implement <tt class="docutils literal"><span class="pre">parse</span></tt> as:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="kt">bool</span> <span class="n">FileSizeParser</span><span class="o">::</span><span class="n">parse</span><span class="p">(</span><span class="n">cl</span><span class="o">::</span><span class="n">Option</span> <span class="o">&</span><span class="n">O</span><span class="p">,</span> <span class="k">const</span> <span class="kt">char</span> <span class="o">*</span><span class="n">ArgName</span><span class="p">,</span>
+                           <span class="k">const</span> <span class="n">std</span><span class="o">::</span><span class="n">string</span> <span class="o">&</span><span class="n">Arg</span><span class="p">,</span> <span class="kt">unsigned</span> <span class="o">&</span><span class="n">Val</span><span class="p">)</span> <span class="p">{</span>
+  <span class="k">const</span> <span class="kt">char</span> <span class="o">*</span><span class="n">ArgStart</span> <span class="o">=</span> <span class="n">Arg</span><span class="p">.</span><span class="n">c_str</span><span class="p">();</span>
+  <span class="kt">char</span> <span class="o">*</span><span class="n">End</span><span class="p">;</span>
+
+  <span class="c1">// Parse integer part, leaving 'End' pointing to the first non-integer char</span>
+  <span class="n">Val</span> <span class="o">=</span> <span class="p">(</span><span class="kt">unsigned</span><span class="p">)</span><span class="n">strtol</span><span class="p">(</span><span class="n">ArgStart</span><span class="p">,</span> <span class="o">&</span><span class="n">End</span><span class="p">,</span> <span class="mi">0</span><span class="p">);</span>
+
+  <span class="k">while</span> <span class="p">(</span><span class="mi">1</span><span class="p">)</span> <span class="p">{</span>
+    <span class="k">switch</span> <span class="p">(</span><span class="o">*</span><span class="n">End</span><span class="o">++</span><span class="p">)</span> <span class="p">{</span>
+    <span class="k">case</span> <span class="mi">0</span>: <span class="k">return</span> <span class="nb">false</span><span class="p">;</span>   <span class="c1">// No error</span>
+    <span class="k">case</span> <span class="sc">'i'</span>:               <span class="c1">// Ignore the 'i' in KiB if people use that</span>
+    <span class="k">case</span> <span class="sc">'b'</span>: <span class="k">case</span> <span class="sc">'B'</span>:     <span class="c1">// Ignore B suffix</span>
+      <span class="k">break</span><span class="p">;</span>
+
+    <span class="k">case</span> <span class="sc">'g'</span>: <span class="k">case</span> <span class="sc">'G'</span>: <span class="n">Val</span> <span class="o">*=</span> <span class="mi">1024</span><span class="o">*</span><span class="mi">1024</span><span class="o">*</span><span class="mi">1024</span><span class="p">;</span> <span class="k">break</span><span class="p">;</span>
+    <span class="k">case</span> <span class="sc">'m'</span>: <span class="k">case</span> <span class="sc">'M'</span>: <span class="n">Val</span> <span class="o">*=</span> <span class="mi">1024</span><span class="o">*</span><span class="mi">1024</span><span class="p">;</span>      <span class="k">break</span><span class="p">;</span>
+    <span class="k">case</span> <span class="sc">'k'</span>: <span class="k">case</span> <span class="sc">'K'</span>: <span class="n">Val</span> <span class="o">*=</span> <span class="mi">1024</span><span class="p">;</span>           <span class="k">break</span><span class="p">;</span>
+
+    <span class="nl">default:</span>
+      <span class="c1">// Print an error message if unrecognized character!</span>
+      <span class="k">return</span> <span class="n">O</span><span class="p">.</span><span class="n">error</span><span class="p">(</span><span class="s">"'"</span> <span class="o">+</span> <span class="n">Arg</span> <span class="o">+</span> <span class="s">"' value invalid for file size argument!"</span><span class="p">);</span>
+    <span class="p">}</span>
+  <span class="p">}</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>This function implements a very simple parser for the kinds of strings we are
+interested in.  Although it has some holes (it allows “<tt class="docutils literal"><span class="pre">123KKK</span></tt>” for example),
+it is good enough for this example.  Note that we use the option itself to print
+out the error message (the <tt class="docutils literal"><span class="pre">error</span></tt> method always returns true) in order to get
+a nice error message (shown below).  Now that we have our parser class, we can
+use it like this:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">static</span> <span class="n">cl</span><span class="o">::</span><span class="n">opt</span><span class="o"><</span><span class="kt">unsigned</span><span class="p">,</span> <span class="nb">false</span><span class="p">,</span> <span class="n">FileSizeParser</span><span class="o">></span>
+<span class="n">MFS</span><span class="p">(</span><span class="s">"max-file-size"</span><span class="p">,</span> <span class="n">cl</span><span class="o">::</span><span class="n">desc</span><span class="p">(</span><span class="s">"Maximum file size to accept"</span><span class="p">),</span>
+    <span class="n">cl</span><span class="o">::</span><span class="n">value_desc</span><span class="p">(</span><span class="s">"size"</span><span class="p">));</span>
+</pre></div>
+</div>
+<p>Which adds this to the output of our program:</p>
+<div class="highlight-python"><pre>OPTIONS:
+  -help                 - display available options (-help-hidden for more)
+  ...
+ -max-file-size=<size> - Maximum file size to accept</pre>
+</div>
+<p>And we can test that our parse works correctly now (the test program just prints
+out the max-file-size argument value):</p>
+<div class="highlight-python"><pre>$ ./test
+MFS: 0
+$ ./test -max-file-size=123MB
+MFS: 128974848
+$ ./test -max-file-size=3G
+MFS: 3221225472
+$ ./test -max-file-size=dog
+-max-file-size option: 'dog' value invalid for file size argument!</pre>
+</div>
+<p>It looks like it works.  The error message that we get is nice and helpful, and
+we seem to accept reasonable file sizes.  This wraps up the “custom parser”
+tutorial.</p>
+</div>
+<div class="section" id="exploiting-external-storage">
+<h3><a class="toc-backref" href="#id46">Exploiting external storage</a><a class="headerlink" href="#exploiting-external-storage" title="Permalink to this headline">¶</a></h3>
+<p>Several of the LLVM libraries define static <tt class="docutils literal"><span class="pre">cl::opt</span></tt> instances that will
+automatically be included in any program that links with that library.  This is
+a feature. However, sometimes it is necessary to know the value of the command
+line option outside of the library. In these cases the library does or should
+provide an external storage location that is accessible to users of the
+library. Examples of this include the <tt class="docutils literal"><span class="pre">llvm::DebugFlag</span></tt> exported by the
+<tt class="docutils literal"><span class="pre">lib/Support/Debug.cpp</span></tt> file and the <tt class="docutils literal"><span class="pre">llvm::TimePassesIsEnabled</span></tt> flag
+exported by the <tt class="docutils literal"><span class="pre">lib/VMCore/PassManager.cpp</span></tt> file.</p>
+<p id="dynamically-loaded-options">Dynamically adding command line options</p>
+</div>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="CompilerWriterInfo.html" title="Architecture & Platform Information for Compiler Writers"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="CodingStandards.html" title="LLVM Coding Standards"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/CompilerWriterInfo.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/CompilerWriterInfo.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/CompilerWriterInfo.html (added)
+++ www-releases/trunk/3.5.1/docs/CompilerWriterInfo.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,300 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>Architecture & Platform Information for Compiler Writers — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="Extending LLVM: Adding instructions, intrinsics, types, etc." href="ExtendingLLVM.html" />
+    <link rel="prev" title="CommandLine 2.0 Library Manual" href="CommandLine.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">
+      <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="ExtendingLLVM.html" title="Extending LLVM: Adding instructions, intrinsics, types, etc."
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="CommandLine.html" title="CommandLine 2.0 Library Manual"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="architecture-platform-information-for-compiler-writers">
+<h1>Architecture & Platform Information for Compiler Writers<a class="headerlink" href="#architecture-platform-information-for-compiler-writers" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#hardware" id="id3">Hardware</a><ul>
+<li><a class="reference internal" href="#arm" id="id4">ARM</a></li>
+<li><a class="reference internal" href="#aarch64" id="id5">AArch64</a></li>
+<li><a class="reference internal" href="#itanium-ia64" id="id6">Itanium (ia64)</a></li>
+<li><a class="reference internal" href="#mips" id="id7">MIPS</a></li>
+<li><a class="reference internal" href="#powerpc" id="id8">PowerPC</a><ul>
+<li><a class="reference internal" href="#ibm-official-manuals-and-docs" id="id9">IBM - Official manuals and docs</a></li>
+<li><a class="reference internal" href="#other-documents-collections-notes" id="id10">Other documents, collections, notes</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#r600" id="id11">R600</a></li>
+<li><a class="reference internal" href="#sparc" id="id12">SPARC</a></li>
+<li><a class="reference internal" href="#systemz" id="id13">SystemZ</a></li>
+<li><a class="reference internal" href="#x86" id="id14">X86</a><ul>
+<li><a class="reference internal" href="#amd-official-manuals-and-docs" id="id15">AMD - Official manuals and docs</a></li>
+<li><a class="reference internal" href="#intel-official-manuals-and-docs" id="id16">Intel - Official manuals and docs</a></li>
+<li><a class="reference internal" href="#other-x86-specific-information" id="id17">Other x86-specific information</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#xcore" id="id18">XCore</a></li>
+<li><a class="reference internal" href="#other-relevant-lists" id="id19">Other relevant lists</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#id2" id="id20">ABI</a><ul>
+<li><a class="reference internal" href="#linux" id="id21">Linux</a></li>
+<li><a class="reference internal" href="#os-x" id="id22">OS X</a></li>
+<li><a class="reference internal" href="#windows" id="id23">Windows</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#nvptx" id="id24">NVPTX</a></li>
+<li><a class="reference internal" href="#miscellaneous-resources" id="id25">Miscellaneous Resources</a></li>
+</ul>
+</div>
+<div class="admonition note">
+<p class="first admonition-title">Note</p>
+<p class="last">This document is a work-in-progress.  Additions and clarifications are
+welcome.</p>
+</div>
+<div class="section" id="hardware">
+<h2><a class="toc-backref" href="#id3">Hardware</a><a class="headerlink" href="#hardware" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="arm">
+<h3><a class="toc-backref" href="#id4">ARM</a><a class="headerlink" href="#arm" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://www.arm.com/documentation/">ARM documentation</a> (<a class="reference external" href="http://www.arm.com/documentation/ARMProcessor_Cores/">Processor Cores</a> Cores)</li>
+<li><a class="reference external" href="http://www.arm.com/products/DevTools/ABI.html">ABI</a></li>
+<li><a class="reference external" href="http://infocenter.arm.com/help/topic/com.arm.doc.ihi0045d/IHI0045D_ABI_addenda.pdf">ABI Addenda and Errata</a></li>
+<li><a class="reference external" href="http://infocenter.arm.com/help/topic/com.arm.doc.ihi0053a/IHI0053A_acle.pdf">ARM C Language Extensions</a></li>
+</ul>
+</div>
+<div class="section" id="aarch64">
+<h3><a class="toc-backref" href="#id5">AArch64</a><a class="headerlink" href="#aarch64" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.genc010197a/index.html">ARMv8 Instruction Set Overview</a></li>
+<li><a class="reference external" href="http://infocenter.arm.com/help/topic/com.arm.doc.ihi0053a/IHI0053A_acle.pdf">ARM C Language Extensions</a></li>
+</ul>
+</div>
+<div class="section" id="itanium-ia64">
+<h3><a class="toc-backref" href="#id6">Itanium (ia64)</a><a class="headerlink" href="#itanium-ia64" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://developer.intel.com/design/itanium2/documentation.htm">Itanium documentation</a></li>
+</ul>
+</div>
+<div class="section" id="mips">
+<h3><a class="toc-backref" href="#id7">MIPS</a><a class="headerlink" href="#mips" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://imgtec.com/mips/mips-architectures.asp">MIPS Processor Architecture</a></li>
+</ul>
+</div>
+<div class="section" id="powerpc">
+<h3><a class="toc-backref" href="#id8">PowerPC</a><a class="headerlink" href="#powerpc" title="Permalink to this headline">¶</a></h3>
+<div class="section" id="ibm-official-manuals-and-docs">
+<h4><a class="toc-backref" href="#id9">IBM - Official manuals and docs</a><a class="headerlink" href="#ibm-official-manuals-and-docs" title="Permalink to this headline">¶</a></h4>
+<ul class="simple">
+<li><a class="reference external" href="https://www.power.org/technology-introduction/standards-specifications">Power Instruction Set Architecture, Versions 2.03 through 2.06 (authentication required, free sign-up)</a></li>
+<li><a class="reference external" href="http://www.ibm.com/chips/techlib/techlib.nsf/techdocs/852569B20050FF7785256996007558C6">PowerPC Compiler Writer’s Guide</a></li>
+<li><a class="reference external" href="http://www.ibm.com/developerworks/linux/library/l-powarch/">Intro to PowerPC Architecture</a></li>
+<li><a class="reference external" href="http://www.ibm.com/chips/techlib/techlib.nsf/products/PowerPC">PowerPC Processor Manuals (embedded)</a></li>
+<li><a class="reference external" href="https://www.power.org/documentation/?document_company=105&document_category=all&publish_year=all&grid_order=DESC&grid_sort=title">Various IBM specifications and white papers</a></li>
+<li><a class="reference external" href="http://publibn.boulder.ibm.com/doc_link/en_US/a_doc_lib/aixassem/alangref/alangreftfrm.htm">IBM AIX/5L for POWER Assembly Reference</a></li>
+</ul>
+</div>
+<div class="section" id="other-documents-collections-notes">
+<h4><a class="toc-backref" href="#id10">Other documents, collections, notes</a><a class="headerlink" href="#other-documents-collections-notes" title="Permalink to this headline">¶</a></h4>
+<ul class="simple">
+<li><a class="reference external" href="http://penguinppc.org/dev/#library">PowerPC ABI documents</a></li>
+<li><a class="reference external" href="http://gcc.gnu.org/ml/gcc-patches/2003-09/msg00997.html">PowerPC64 alignment of long doubles (from GCC)</a></li>
+<li><a class="reference external" href="http://sources.redhat.com/ml/binutils/2002-04/msg00573.html">Long branch stubs for powerpc64-linux (from binutils)</a></li>
+</ul>
+</div>
+</div>
+<div class="section" id="r600">
+<h3><a class="toc-backref" href="#id11">R600</a><a class="headerlink" href="#r600" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://developer.amd.com/wordpress/media/2012/10/R600_Instruction_Set_Architecture.pdf">AMD R6xx shader ISA</a></li>
+<li><a class="reference external" href="http://developer.amd.com/wordpress/media/2012/10/R700-Family_Instruction_Set_Architecture.pdf">AMD R7xx shader ISA</a></li>
+<li><a class="reference external" href="http://developer.amd.com/wordpress/media/2012/10/AMD_Evergreen-Family_Instruction_Set_Architecture.pdf">AMD Evergreen shader ISA</a></li>
+<li><a class="reference external" href="http://developer.amd.com/wordpress/media/2012/10/AMD_HD_6900_Series_Instruction_Set_Architecture.pdf">AMD Cayman/Trinity shader ISA</a></li>
+<li><a class="reference external" href="http://developer.amd.com/wordpress/media/2012/12/AMD_Southern_Islands_Instruction_Set_Architecture.pdf">AMD Southern Islands Series ISA</a></li>
+<li><a class="reference external" href="http://developer.amd.com/download/AMD_Accelerated_Parallel_Processing_OpenCL_Programming_Guide.pdf">AMD GPU Programming Guide</a></li>
+<li><a class="reference external" href="http://developer.amd.com/tools/heterogeneous-computing/amd-accelerated-parallel-processing-app-sdk/documentation/">AMD Compute Resources</a></li>
+</ul>
+</div>
+<div class="section" id="sparc">
+<h3><a class="toc-backref" href="#id12">SPARC</a><a class="headerlink" href="#sparc" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://sparc.org/standards">SPARC standards</a></li>
+<li><a class="reference external" href="http://sparc.org/standards/64.psabi.1.35.ps.Z">SPARC V9 ABI</a></li>
+<li><a class="reference external" href="http://sparc.org/standards/psABI3rd.pdf">SPARC V8 ABI</a></li>
+</ul>
+</div>
+<div class="section" id="systemz">
+<h3><a class="toc-backref" href="#id13">SystemZ</a><a class="headerlink" href="#systemz" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://www-01.ibm.com/support/docview.wss?uid=isg2b9de5f05a9d57819852571c500428f9a">z/Architecture Principles of Operation (registration required, free sign-up)</a></li>
+</ul>
+</div>
+<div class="section" id="x86">
+<h3><a class="toc-backref" href="#id14">X86</a><a class="headerlink" href="#x86" title="Permalink to this headline">¶</a></h3>
+<div class="section" id="amd-official-manuals-and-docs">
+<h4><a class="toc-backref" href="#id15">AMD - Official manuals and docs</a><a class="headerlink" href="#amd-official-manuals-and-docs" title="Permalink to this headline">¶</a></h4>
+<ul class="simple">
+<li><a class="reference external" href="http://www.amd.com/us-en/Processors/TechnicalResources/0,,30_182_739,00.html">AMD processor manuals</a></li>
+<li><a class="reference external" href="http://www.x86-64.org/documentation">X86-64 ABI</a></li>
+</ul>
+</div>
+<div class="section" id="intel-official-manuals-and-docs">
+<h4><a class="toc-backref" href="#id16">Intel - Official manuals and docs</a><a class="headerlink" href="#intel-official-manuals-and-docs" title="Permalink to this headline">¶</a></h4>
+<ul class="simple">
+<li><a class="reference external" href="http://www.intel.com/content/www/us/en/processors/architectures-software-developer-manuals.html">Intel 64 and IA-32 manuals</a></li>
+<li><a class="reference external" href="http://www.intel.com/design/itanium/documentation.htm?iid=ipp_srvr_proc_itanium2+techdocs">Intel Itanium documentation</a></li>
+</ul>
+</div>
+<div class="section" id="other-x86-specific-information">
+<h4><a class="toc-backref" href="#id17">Other x86-specific information</a><a class="headerlink" href="#other-x86-specific-information" title="Permalink to this headline">¶</a></h4>
+<ul class="simple">
+<li><a class="reference external" href="http://www.agner.org/optimize/calling_conventions.pdf">Calling conventions for different C++ compilers and operating systems</a></li>
+</ul>
+</div>
+</div>
+<div class="section" id="xcore">
+<h3><a class="toc-backref" href="#id18">XCore</a><a class="headerlink" href="#xcore" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="https://www.xmos.com/en/download/public/The-XMOS-XS1-Architecture%28X7879A%29.pdf">The XMOS XS1 Architecture (ISA)</a></li>
+<li><a class="reference external" href="https://www.xmos.com/download/public/Tools-Development-Guide%28X9114A%29.pdf">Tools Development Guide (includes ABI)</a></li>
+</ul>
+</div>
+<div class="section" id="other-relevant-lists">
+<h3><a class="toc-backref" href="#id19">Other relevant lists</a><a class="headerlink" href="#other-relevant-lists" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://gcc.gnu.org/readings.html">GCC reading list</a></li>
+</ul>
+</div>
+</div>
+<div class="section" id="id2">
+<h2><a class="toc-backref" href="#id20">ABI</a><a class="headerlink" href="#id2" title="Permalink to this headline">¶</a></h2>
+<ul class="simple">
+<li><a class="reference external" href="http://www.sco.com/developers/gabi/latest/contents.html">System V Application Binary Interface</a></li>
+<li><a class="reference external" href="http://mentorembedded.github.io/cxx-abi/">Itanium C++ ABI</a></li>
+</ul>
+<div class="section" id="linux">
+<h3><a class="toc-backref" href="#id21">Linux</a><a class="headerlink" href="#linux" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://www.linuxbase.org/spec/ELF/ppc64/">PowerPC 64-bit ELF ABI Supplement</a></li>
+<li><a class="reference external" href="http://infocenter.arm.com/help/topic/com.arm.doc.ihi0055a/IHI0055A_aapcs64.pdf">Procedure Call Standard for the AArch64 Architecture</a></li>
+<li><a class="reference external" href="http://infocenter.arm.com/help/topic/com.arm.doc.ihi0044e/IHI0044E_aaelf.pdf">ELF for the ARM Architecture</a></li>
+<li><a class="reference external" href="http://infocenter.arm.com/help/topic/com.arm.doc.ihi0056a/IHI0056A_aaelf64.pdf">ELF for the ARM 64-bit Architecture (AArch64)</a></li>
+<li><a class="reference external" href="http://legacy.redhat.com/pub/redhat/linux/7.1/es/os/s390x/doc/lzsabi0.pdf">System z ELF ABI Supplement</a></li>
+</ul>
+</div>
+<div class="section" id="os-x">
+<h3><a class="toc-backref" href="#id22">OS X</a><a class="headerlink" href="#os-x" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://developer.apple.com/documentation/Darwin/RuntimeArchitecture-date.html">Mach-O Runtime Architecture</a></li>
+<li><a class="reference external" href="http://www.unsanity.org/archives/000044.php">Notes on Mach-O ABI</a></li>
+</ul>
+</div>
+<div class="section" id="windows">
+<h3><a class="toc-backref" href="#id23">Windows</a><a class="headerlink" href="#windows" title="Permalink to this headline">¶</a></h3>
+<ul class="simple">
+<li><a class="reference external" href="http://www.microsoft.com/whdc/system/platform/firmware/pecoff.mspx">Microsoft PE/COFF Specification</a></li>
+</ul>
+</div>
+</div>
+<div class="section" id="nvptx">
+<h2><a class="toc-backref" href="#id24">NVPTX</a><a class="headerlink" href="#nvptx" title="Permalink to this headline">¶</a></h2>
+<ul class="simple">
+<li><a class="reference external" href="http://docs.nvidia.com/cuda/index.html">CUDA Documentation</a> includes the PTX
+ISA and Driver API documentation</li>
+</ul>
+</div>
+<div class="section" id="miscellaneous-resources">
+<h2><a class="toc-backref" href="#id25">Miscellaneous Resources</a><a class="headerlink" href="#miscellaneous-resources" title="Permalink to this headline">¶</a></h2>
+<ul class="simple">
+<li><a class="reference external" href="http://www.nondot.org/sabre/os/articles/ExecutableFileFormats/">Executable File Format library</a></li>
+<li><a class="reference external" href="http://gcc.gnu.org/projects/prefetch.html">GCC prefetch project</a> page has a
+good survey of the prefetching capabilities of a variety of modern
+processors.</li>
+</ul>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="ExtendingLLVM.html" title="Extending LLVM: Adding instructions, intrinsics, types, etc."
+             >next</a> |</li>
+        <li class="right" >
+          <a href="CommandLine.html" title="CommandLine 2.0 Library Manual"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/DebuggingJITedCode.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/DebuggingJITedCode.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/DebuggingJITedCode.html (added)
+++ www-releases/trunk/3.5.1/docs/DebuggingJITedCode.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,222 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>Debugging JIT-ed Code With GDB — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="The LLVM gold plugin" href="GoldPlugin.html" />
+    <link rel="prev" title="TableGen Deficiencies" href="TableGen/Deficiencies.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">
+      <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="GoldPlugin.html" title="The LLVM gold plugin"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="TableGen/Deficiencies.html" title="TableGen Deficiencies"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="debugging-jit-ed-code-with-gdb">
+<h1>Debugging JIT-ed Code With GDB<a class="headerlink" href="#debugging-jit-ed-code-with-gdb" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="background">
+<h2>Background<a class="headerlink" href="#background" title="Permalink to this headline">¶</a></h2>
+<p>Without special runtime support, debugging dynamically generated code with
+GDB (as well as most debuggers) can be quite painful.  Debuggers generally
+read debug information from the object file of the code, but for JITed
+code, there is no such file to look for.</p>
+<p>In order to communicate the necessary debug info to GDB, an interface for
+registering JITed code with debuggers has been designed and implemented for
+GDB and LLVM MCJIT.  At a high level, whenever MCJIT generates new machine code,
+it does so in an in-memory object file that contains the debug information in
+DWARF format.  MCJIT then adds this in-memory object file to a global list of
+dynamically generated object files and calls a special function
+(<tt class="docutils literal"><span class="pre">__jit_debug_register_code</span></tt>) marked noinline that GDB knows about.  When
+GDB attaches to a process, it puts a breakpoint in this function and loads all
+of the object files in the global list.  When MCJIT calls the registration
+function, GDB catches the breakpoint signal, loads the new object file from
+the inferior’s memory, and resumes the execution.  In this way, GDB can get the
+necessary debug information.</p>
+</div>
+<div class="section" id="gdb-version">
+<h2>GDB Version<a class="headerlink" href="#gdb-version" title="Permalink to this headline">¶</a></h2>
+<p>In order to debug code JIT-ed by LLVM, you need GDB 7.0 or newer, which is
+available on most modern distributions of Linux.  The version of GDB that
+Apple ships with Xcode has been frozen at 6.3 for a while.  LLDB may be a
+better option for debugging JIT-ed code on Mac OS X.</p>
+</div>
+<div class="section" id="debugging-mcjit-ed-code">
+<h2>Debugging MCJIT-ed code<a class="headerlink" href="#debugging-mcjit-ed-code" title="Permalink to this headline">¶</a></h2>
+<p>The emerging MCJIT component of LLVM allows full debugging of JIT-ed code with
+GDB.  This is due to MCJIT’s ability to use the MC emitter to provide full
+DWARF debugging information to GDB.</p>
+<p>Note that lli has to be passed the <tt class="docutils literal"><span class="pre">-use-mcjit</span></tt> flag to JIT the code with
+MCJIT instead of the old JIT.</p>
+<div class="section" id="example">
+<h3>Example<a class="headerlink" href="#example" title="Permalink to this headline">¶</a></h3>
+<p>Consider the following C code (with line numbers added to make the example
+easier to follow):</p>
+<div class="highlight-c"><div class="highlight"><pre><span class="mi">1</span>   <span class="kt">int</span> <span class="n">compute_factorial</span><span class="p">(</span><span class="kt">int</span> <span class="n">n</span><span class="p">)</span>
+<span class="mi">2</span>   <span class="p">{</span>
+<span class="mi">3</span>       <span class="k">if</span> <span class="p">(</span><span class="n">n</span> <span class="o"><=</span> <span class="mi">1</span><span class="p">)</span>
+<span class="mi">4</span>           <span class="k">return</span> <span class="mi">1</span><span class="p">;</span>
+<span class="mi">5</span>
+<span class="mi">6</span>       <span class="kt">int</span> <span class="n">f</span> <span class="o">=</span> <span class="n">n</span><span class="p">;</span>
+<span class="mi">7</span>       <span class="k">while</span> <span class="p">(</span><span class="o">--</span><span class="n">n</span> <span class="o">></span> <span class="mi">1</span><span class="p">)</span>
+<span class="mi">8</span>           <span class="n">f</span> <span class="o">*=</span> <span class="n">n</span><span class="p">;</span>
+<span class="mi">9</span>       <span class="k">return</span> <span class="n">f</span><span class="p">;</span>
+<span class="mi">10</span>  <span class="p">}</span>
+<span class="mi">11</span>
+<span class="mi">12</span>
+<span class="mi">13</span>  <span class="kt">int</span> <span class="n">main</span><span class="p">(</span><span class="kt">int</span> <span class="n">argc</span><span class="p">,</span> <span class="kt">char</span><span class="o">**</span> <span class="n">argv</span><span class="p">)</span>
+<span class="mi">14</span>  <span class="p">{</span>
+<span class="mi">15</span>      <span class="k">if</span> <span class="p">(</span><span class="n">argc</span> <span class="o"><</span> <span class="mi">2</span><span class="p">)</span>
+<span class="mi">16</span>          <span class="k">return</span> <span class="o">-</span><span class="mi">1</span><span class="p">;</span>
+<span class="mi">17</span>      <span class="kt">char</span> <span class="n">firstletter</span> <span class="o">=</span> <span class="n">argv</span><span class="p">[</span><span class="mi">1</span><span class="p">][</span><span class="mi">0</span><span class="p">];</span>
+<span class="mi">18</span>      <span class="kt">int</span> <span class="n">result</span> <span class="o">=</span> <span class="n">compute_factorial</span><span class="p">(</span><span class="n">firstletter</span> <span class="o">-</span> <span class="sc">'0'</span><span class="p">);</span>
+<span class="mi">19</span>
+<span class="mi">20</span>      <span class="c1">// Returned result is clipped at 255...</span>
+<span class="mi">21</span>      <span class="k">return</span> <span class="n">result</span><span class="p">;</span>
+<span class="mi">22</span>  <span class="p">}</span>
+</pre></div>
+</div>
+<p>Here is a sample command line session that shows how to build and run this
+code via <tt class="docutils literal"><span class="pre">lli</span></tt> inside GDB:</p>
+<div class="highlight-bash"><div class="highlight"><pre><span class="nv">$ $BINPATH</span>/clang -cc1 -O0 -g -emit-llvm showdebug.c
+<span class="nv">$ </span>gdb --quiet --args <span class="nv">$BINPATH</span>/lli -use-mcjit showdebug.ll 5
+Reading symbols from <span class="nv">$BINPATH</span>/lli...done.
+<span class="o">(</span>gdb<span class="o">)</span> b showdebug.c:6
+No <span class="nb">source </span>file named showdebug.c.
+Make breakpoint pending on future shared library load? <span class="o">(</span>y or <span class="o">[</span>n<span class="o">])</span> y
+Breakpoint 1 <span class="o">(</span>showdebug.c:6<span class="o">)</span> pending.
+<span class="o">(</span>gdb<span class="o">)</span> r
+Starting program: <span class="nv">$BINPATH</span>/lli -use-mcjit showdebug.ll 5
+<span class="o">[</span>Thread debugging using libthread_db enabled<span class="o">]</span>
+
+Breakpoint 1, compute_factorial <span class="o">(</span><span class="nv">n</span><span class="o">=</span>5<span class="o">)</span> at showdebug.c:6
+6        int <span class="nv">f</span> <span class="o">=</span> n;
+<span class="o">(</span>gdb<span class="o">)</span> p n
+<span class="nv">$1</span> <span class="o">=</span> 5
+<span class="o">(</span>gdb<span class="o">)</span> p f
+<span class="nv">$2</span> <span class="o">=</span> 0
+<span class="o">(</span>gdb<span class="o">)</span> n
+7        <span class="k">while</span> <span class="o">(</span>--n > 1<span class="o">)</span>
+<span class="o">(</span>gdb<span class="o">)</span> p f
+<span class="nv">$3</span> <span class="o">=</span> 5
+<span class="o">(</span>gdb<span class="o">)</span> b showdebug.c:9
+Breakpoint 2 at 0x7ffff7ed404c: file showdebug.c, line 9.
+<span class="o">(</span>gdb<span class="o">)</span> c
+Continuing.
+
+Breakpoint 2, compute_factorial <span class="o">(</span><span class="nv">n</span><span class="o">=</span>1<span class="o">)</span> at showdebug.c:9
+9        <span class="k">return </span>f;
+<span class="o">(</span>gdb<span class="o">)</span> p f
+<span class="nv">$4</span> <span class="o">=</span> 120
+<span class="o">(</span>gdb<span class="o">)</span> bt
+<span class="c">#0  compute_factorial (n=1) at showdebug.c:9</span>
+<span class="c">#1  0x00007ffff7ed40a9 in main (argc=2, argv=0x16677e0) at showdebug.c:18</span>
+<span class="c">#2  0x3500000001652748 in ?? ()</span>
+<span class="c">#3  0x00000000016677e0 in ?? ()</span>
+<span class="c">#4  0x0000000000000002 in ?? ()</span>
+<span class="c">#5  0x0000000000d953b3 in llvm::MCJIT::runFunction (this=0x16151f0, F=0x1603020, ArgValues=...) at /home/ebenders_test/llvm_svn_rw/lib/ExecutionEngine/MCJIT/MCJIT.cpp:161</span>
+<span class="c">#6  0x0000000000dc8872 in llvm::ExecutionEngine::runFunctionAsMain (this=0x16151f0, Fn=0x1603020, argv=..., envp=0x7fffffffe040)</span>
+    at /home/ebenders_test/llvm_svn_rw/lib/ExecutionEngine/ExecutionEngine.cpp:397
+<span class="c">#7  0x000000000059c583 in main (argc=4, argv=0x7fffffffe018, envp=0x7fffffffe040) at /home/ebenders_test/llvm_svn_rw/tools/lli/lli.cpp:324</span>
+<span class="o">(</span>gdb<span class="o">)</span> finish
+Run till <span class="nb">exit </span>from <span class="c">#0  compute_factorial (n=1) at showdebug.c:9</span>
+0x00007ffff7ed40a9 in main <span class="o">(</span><span class="nv">argc</span><span class="o">=</span>2, <span class="nv">argv</span><span class="o">=</span>0x16677e0<span class="o">)</span> at showdebug.c:18
+18       int <span class="nv">result</span> <span class="o">=</span> compute_factorial<span class="o">(</span>firstletter - <span class="s1">'0'</span><span class="o">)</span>;
+Value returned is <span class="nv">$5</span> <span class="o">=</span> 120
+<span class="o">(</span>gdb<span class="o">)</span> p result
+<span class="nv">$6</span> <span class="o">=</span> 23406408
+<span class="o">(</span>gdb<span class="o">)</span> n
+21       <span class="k">return </span>result;
+<span class="o">(</span>gdb<span class="o">)</span> p result
+<span class="nv">$7</span> <span class="o">=</span> 120
+<span class="o">(</span>gdb<span class="o">)</span> c
+Continuing.
+
+Program exited with code 0170.
+<span class="o">(</span>gdb<span class="o">)</span>
+</pre></div>
+</div>
+</div>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="GoldPlugin.html" title="The LLVM gold plugin"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="TableGen/Deficiencies.html" title="TableGen Deficiencies"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/DeveloperPolicy.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/DeveloperPolicy.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/DeveloperPolicy.html (added)
+++ www-releases/trunk/3.5.1/docs/DeveloperPolicy.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,563 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>LLVM Developer Policy — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="LLVM Makefile Guide" href="MakefileGuide.html" />
+    <link rel="prev" title="Using ARM NEON instructions in big endian mode" href="BigEndianNEON.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">
+      <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="MakefileGuide.html" title="LLVM Makefile Guide"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="BigEndianNEON.html" title="Using ARM NEON instructions in big endian mode"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-developer-policy">
+<h1>LLVM Developer Policy<a class="headerlink" href="#llvm-developer-policy" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#introduction" id="id6">Introduction</a></li>
+<li><a class="reference internal" href="#developer-policies" id="id7">Developer Policies</a><ul>
+<li><a class="reference internal" href="#stay-informed" id="id8">Stay Informed</a></li>
+<li><a class="reference internal" href="#making-and-submitting-a-patch" id="id9">Making and Submitting a Patch</a></li>
+<li><a class="reference internal" href="#code-reviews" id="id10">Code Reviews</a></li>
+<li><a class="reference internal" href="#code-owners" id="id11">Code Owners</a></li>
+<li><a class="reference internal" href="#test-cases" id="id12">Test Cases</a></li>
+<li><a class="reference internal" href="#quality" id="id13">Quality</a></li>
+<li><a class="reference internal" href="#obtaining-commit-access" id="id14">Obtaining Commit Access</a></li>
+<li><a class="reference internal" href="#making-a-major-change" id="id15">Making a Major Change</a></li>
+<li><a class="reference internal" href="#incremental-development" id="id16">Incremental Development</a></li>
+<li><a class="reference internal" href="#attribution-of-changes" id="id17">Attribution of Changes</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#copyright-license-and-patents" id="id18">Copyright, License, and Patents</a><ul>
+<li><a class="reference internal" href="#copyright" id="id19">Copyright</a></li>
+<li><a class="reference internal" href="#license" id="id20">License</a></li>
+<li><a class="reference internal" href="#patents" id="id21">Patents</a></li>
+</ul>
+</li>
+</ul>
+</div>
+<div class="section" id="introduction">
+<h2><a class="toc-backref" href="#id6">Introduction</a><a class="headerlink" href="#introduction" title="Permalink to this headline">¶</a></h2>
+<p>This document contains the LLVM Developer Policy which defines the project’s
+policy towards developers and their contributions. The intent of this policy is
+to eliminate miscommunication, rework, and confusion that might arise from the
+distributed nature of LLVM’s development.  By stating the policy in clear terms,
+we hope each developer can know ahead of time what to expect when making LLVM
+contributions.  This policy covers all llvm.org subprojects, including Clang,
+LLDB, libc++, etc.</p>
+<p>This policy is also designed to accomplish the following objectives:</p>
+<ol class="arabic simple">
+<li>Attract both users and developers to the LLVM project.</li>
+<li>Make life as simple and easy for contributors as possible.</li>
+<li>Keep the top of Subversion trees as stable as possible.</li>
+<li>Establish awareness of the project’s <a class="reference internal" href="#copyright-license-patents"><em>copyright, license, and patent
+policies</em></a> with contributors to the project.</li>
+</ol>
+<p>This policy is aimed at frequent contributors to LLVM. People interested in
+contributing one-off patches can do so in an informal way by sending them to the
+<a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">llvm-commits mailing list</a> and engaging another
+developer to see it through the process.</p>
+</div>
+<div class="section" id="developer-policies">
+<h2><a class="toc-backref" href="#id7">Developer Policies</a><a class="headerlink" href="#developer-policies" title="Permalink to this headline">¶</a></h2>
+<p>This section contains policies that pertain to frequent LLVM developers.  We
+always welcome <a class="reference internal" href="#one-off-patches">one-off patches</a> from people who do not routinely contribute to
+LLVM, but we expect more from frequent contributors to keep the system as
+efficient as possible for everyone.  Frequent LLVM contributors are expected to
+meet the following requirements in order for LLVM to maintain a high standard of
+quality.</p>
+<div class="section" id="stay-informed">
+<h3><a class="toc-backref" href="#id8">Stay Informed</a><a class="headerlink" href="#stay-informed" title="Permalink to this headline">¶</a></h3>
+<p>Developers should stay informed by reading at least the “dev” mailing list for
+the projects you are interested in, such as <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev</a> for LLVM, <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev">cfe-dev</a> for Clang, or <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev">lldb-dev</a> for LLDB.  If you are
+doing anything more than just casual work on LLVM, it is suggested that you also
+subscribe to the “commits” mailing list for the subproject you’re interested in,
+such as <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">llvm-commits</a>, <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits">cfe-commits</a>, or <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits">lldb-commits</a>.  Reading the
+“commits” list and paying attention to changes being made by others is a good
+way to see what other people are interested in and watching the flow of the
+project as a whole.</p>
+<p>We recommend that active developers register an email account with <a class="reference external" href="http://llvm.org/bugs/">LLVM
+Bugzilla</a> and preferably subscribe to the <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs">llvm-bugs</a> email list to keep track
+of bugs and enhancements occurring in LLVM.  We really appreciate people who are
+proactive at catching incoming bugs in their components and dealing with them
+promptly.</p>
+<p>Please be aware that all public LLVM mailing lists are public and archived, and
+that notices of confidentiality or non-disclosure cannot be respected.</p>
+</div>
+<div class="section" id="making-and-submitting-a-patch">
+<span id="one-off-patches"></span><span id="patch"></span><h3><a class="toc-backref" href="#id9">Making and Submitting a Patch</a><a class="headerlink" href="#making-and-submitting-a-patch" title="Permalink to this headline">¶</a></h3>
+<p>When making a patch for review, the goal is to make it as easy for the reviewer
+to read it as possible.  As such, we recommend that you:</p>
+<ol class="arabic simple">
+<li>Make your patch against the Subversion trunk, not a branch, and not an old
+version of LLVM.  This makes it easy to apply the patch.  For information on
+how to check out SVN trunk, please see the <a class="reference external" href="GettingStarted.html#checkout">Getting Started
+Guide</a>.</li>
+<li>Similarly, patches should be submitted soon after they are generated.  Old
+patches may not apply correctly if the underlying code changes between the
+time the patch was created and the time it is applied.</li>
+<li>Patches should be made with <tt class="docutils literal"><span class="pre">svn</span> <span class="pre">diff</span></tt>, or similar. If you use a
+different tool, make sure it uses the <tt class="docutils literal"><span class="pre">diff</span> <span class="pre">-u</span></tt> format and that it
+doesn’t contain clutter which makes it hard to read.</li>
+<li>If you are modifying generated files, such as the top-level <tt class="docutils literal"><span class="pre">configure</span></tt>
+script, please separate out those changes into a separate patch from the rest
+of your changes.</li>
+</ol>
+<p>Once your patch is ready, submit it by emailing it to the appropriate project’s
+commit mailing list (or commit it directly if applicable). Alternatively, some
+patches get sent to the project’s development list or component of the LLVM bug
+tracker, but the commit list is the primary place for reviews and should
+generally be preferred.</p>
+<p>When sending a patch to a mailing list, it is a good idea to send it as an
+<em>attachment</em> to the message, not embedded into the text of the message.  This
+ensures that your mailer will not mangle the patch when it sends it (e.g. by
+making whitespace changes or by wrapping lines).</p>
+<p><em>For Thunderbird users:</em> Before submitting a patch, please open <em>Preferences >
+Advanced > General > Config Editor</em>, find the key
+<tt class="docutils literal"><span class="pre">mail.content_disposition_type</span></tt>, and set its value to <tt class="docutils literal"><span class="pre">1</span></tt>. Without this
+setting, Thunderbird sends your attachment using <tt class="docutils literal"><span class="pre">Content-Disposition:</span> <span class="pre">inline</span></tt>
+rather than <tt class="docutils literal"><span class="pre">Content-Disposition:</span> <span class="pre">attachment</span></tt>. Apple Mail gamely displays such
+a file inline, making it difficult to work with for reviewers using that
+program.</p>
+<p>When submitting patches, please do not add confidentiality or non-disclosure
+notices to the patches themselves.  These notices conflict with the <a class="reference internal" href="#llvm-license">LLVM
+License</a> and may result in your contribution being excluded.</p>
+</div>
+<div class="section" id="code-reviews">
+<span id="code-review"></span><h3><a class="toc-backref" href="#id10">Code Reviews</a><a class="headerlink" href="#code-reviews" title="Permalink to this headline">¶</a></h3>
+<p>LLVM has a code review policy. Code review is one way to increase the quality of
+software. We generally follow these policies:</p>
+<ol class="arabic simple">
+<li>All developers are required to have significant changes reviewed before they
+are committed to the repository.</li>
+<li>Code reviews are conducted by email on the relevant project’s commit mailing
+list, or alternatively on the project’s development list or bug tracker.</li>
+<li>Code can be reviewed either before it is committed or after.  We expect major
+changes to be reviewed before being committed, but smaller changes (or
+changes where the developer owns the component) can be reviewed after commit.</li>
+<li>The developer responsible for a code change is also responsible for making
+all necessary review-related changes.</li>
+<li>Code review can be an iterative process, which continues until the patch is
+ready to be committed. Specifically, once a patch is sent out for review, it
+needs an explicit “looks good” before it is submitted. Do not assume silent
+approval, or request active objections to the patch with a deadline.</li>
+</ol>
+<p>Sometimes code reviews will take longer than you would hope for, especially for
+larger features. Accepted ways to speed up review times for your patches are:</p>
+<ul class="simple">
+<li>Review other people’s patches. If you help out, everybody will be more
+willing to do the same for you; goodwill is our currency.</li>
+<li>Ping the patch. If it is urgent, provide reasons why it is important to you to
+get this patch landed and ping it every couple of days. If it is
+not urgent, the common courtesy ping rate is one week. Remember that you’re
+asking for valuable time from other professional developers.</li>
+<li>Ask for help on IRC. Developers on IRC will be able to either help you
+directly, or tell you who might be a good reviewer.</li>
+<li>Split your patch into multiple smaller patches that build on each other. The
+smaller your patch, the higher the probability that somebody will take a quick
+look at it.</li>
+</ul>
+<p>Developers should participate in code reviews as both reviewers and
+reviewees. If someone is kind enough to review your code, you should return the
+favor for someone else.  Note that anyone is welcome to review and give feedback
+on a patch, but only people with Subversion write access can approve it.</p>
+<p>There is a web based code review tool that can optionally be used
+for code reviews. See <a class="reference internal" href="Phabricator.html"><em>Code Reviews with Phabricator</em></a>.</p>
+</div>
+<div class="section" id="code-owners">
+<h3><a class="toc-backref" href="#id11">Code Owners</a><a class="headerlink" href="#code-owners" title="Permalink to this headline">¶</a></h3>
+<p>The LLVM Project relies on two features of its process to maintain rapid
+development in addition to the high quality of its source base: the combination
+of code review plus post-commit review for trusted maintainers.  Having both is
+a great way for the project to take advantage of the fact that most people do
+the right thing most of the time, and only commit patches without pre-commit
+review when they are confident they are right.</p>
+<p>The trick to this is that the project has to guarantee that all patches that are
+committed are reviewed after they go in: you don’t want everyone to assume
+someone else will review it, allowing the patch to go unreviewed.  To solve this
+problem, we have a notion of an ‘owner’ for a piece of the code.  The sole
+responsibility of a code owner is to ensure that a commit to their area of the
+code is appropriately reviewed, either by themself or by someone else.  The list
+of current code owners can be found in the file
+<a class="reference external" href="http://llvm.org/viewvc/llvm-project/llvm/trunk/CODE_OWNERS.TXT?view=markup">CODE_OWNERS.TXT</a>
+in the root of the LLVM source tree.</p>
+<p>Note that code ownership is completely different than reviewers: anyone can
+review a piece of code, and we welcome code review from anyone who is
+interested.  Code owners are the “last line of defense” to guarantee that all
+patches that are committed are actually reviewed.</p>
+<p>Being a code owner is a somewhat unglamorous position, but it is incredibly
+important for the ongoing success of the project.  Because people get busy,
+interests change, and unexpected things happen, code ownership is purely opt-in,
+and anyone can choose to resign their “title” at any time. For now, we do not
+have an official policy on how one gets elected to be a code owner.</p>
+</div>
+<div class="section" id="test-cases">
+<span id="include-a-testcase"></span><h3><a class="toc-backref" href="#id12">Test Cases</a><a class="headerlink" href="#test-cases" title="Permalink to this headline">¶</a></h3>
+<p>Developers are required to create test cases for any bugs fixed and any new
+features added.  Some tips for getting your testcase approved:</p>
+<ul class="simple">
+<li>All feature and regression test cases are added to the <tt class="docutils literal"><span class="pre">llvm/test</span></tt>
+directory. The appropriate sub-directory should be selected (see the
+<a class="reference internal" href="TestingGuide.html"><em>Testing Guide</em></a> for details).</li>
+<li>Test cases should be written in <a class="reference internal" href="LangRef.html"><em>LLVM assembly language</em></a>.</li>
+<li>Test cases, especially for regressions, should be reduced as much as possible,
+by <a class="reference internal" href="Bugpoint.html"><em>bugpoint</em></a> or manually. It is unacceptable to place an
+entire failing program into <tt class="docutils literal"><span class="pre">llvm/test</span></tt> as this creates a <em>time-to-test</em>
+burden on all developers. Please keep them short.</li>
+</ul>
+<p>Note that llvm/test and clang/test are designed for regression and small feature
+tests only. More extensive test cases (e.g., entire applications, benchmarks,
+etc) should be added to the <tt class="docutils literal"><span class="pre">llvm-test</span></tt> test suite.  The llvm-test suite is
+for coverage (correctness, performance, etc) testing, not feature or regression
+testing.</p>
+</div>
+<div class="section" id="quality">
+<h3><a class="toc-backref" href="#id13">Quality</a><a class="headerlink" href="#quality" title="Permalink to this headline">¶</a></h3>
+<p>The minimum quality standards that any change must satisfy before being
+committed to the main development branch are:</p>
+<ol class="arabic simple">
+<li>Code must adhere to the <a class="reference external" href="CodingStandards.html">LLVM Coding Standards</a>.</li>
+<li>Code must compile cleanly (no errors, no warnings) on at least one platform.</li>
+<li>Bug fixes and new features should <a class="reference internal" href="#include-a-testcase">include a testcase</a> so we know if the
+fix/feature ever regresses in the future.</li>
+<li>Code must pass the <tt class="docutils literal"><span class="pre">llvm/test</span></tt> test suite.</li>
+<li>The code must not cause regressions on a reasonable subset of llvm-test,
+where “reasonable” depends on the contributor’s judgement and the scope of
+the change (more invasive changes require more testing). A reasonable subset
+might be something like “<tt class="docutils literal"><span class="pre">llvm-test/MultiSource/Benchmarks</span></tt>”.</li>
+</ol>
+<p>Additionally, the committer is responsible for addressing any problems found in
+the future that the change is responsible for.  For example:</p>
+<ul class="simple">
+<li>The code should compile cleanly on all supported platforms.</li>
+<li>The changes should not cause any correctness regressions in the <tt class="docutils literal"><span class="pre">llvm-test</span></tt>
+suite and must not cause any major performance regressions.</li>
+<li>The change set should not cause performance or correctness regressions for the
+LLVM tools.</li>
+<li>The changes should not cause performance or correctness regressions in code
+compiled by LLVM on all applicable targets.</li>
+<li>You are expected to address any <a class="reference external" href="http://llvm.org/bugs/">Bugzilla bugs</a> that
+result from your change.</li>
+</ul>
+<p>We prefer for this to be handled before submission but understand that it isn’t
+possible to test all of this for every submission.  Our build bots and nightly
+testing infrastructure normally finds these problems.  A good rule of thumb is
+to check the nightly testers for regressions the day after your change.  Build
+bots will directly email you if a group of commits that included yours caused a
+failure.  You are expected to check the build bot messages to see if they are
+your fault and, if so, fix the breakage.</p>
+<p>Commits that violate these quality standards (e.g. are very broken) may be
+reverted. This is necessary when the change blocks other developers from making
+progress. The developer is welcome to re-commit the change after the problem has
+been fixed.</p>
+</div>
+<div class="section" id="obtaining-commit-access">
+<h3><a class="toc-backref" href="#id14">Obtaining Commit Access</a><a class="headerlink" href="#obtaining-commit-access" title="Permalink to this headline">¶</a></h3>
+<p>We grant commit access to contributors with a track record of submitting high
+quality patches.  If you would like commit access, please send an email to
+<a class="reference external" href="mailto:sabre%40nondot.org">Chris</a> with the following information:</p>
+<ol class="arabic simple">
+<li>The user name you want to commit with, e.g. “hacker”.</li>
+<li>The full name and email address you want message to llvm-commits to come
+from, e.g. “J. Random Hacker <<a class="reference external" href="mailto:hacker%40yoyodyne.com">hacker<span>@</span>yoyodyne<span>.</span>com</a>>”.</li>
+<li>A “password hash” of the password you want to use, e.g. “<tt class="docutils literal"><span class="pre">2ACR96qjUqsyM</span></tt>”.
+Note that you don’t ever tell us what your password is; you just give it to
+us in an encrypted form.  To get this, run “<tt class="docutils literal"><span class="pre">htpasswd</span></tt>” (a utility that
+comes with apache) in crypt mode (often enabled with “<tt class="docutils literal"><span class="pre">-d</span></tt>”), or find a web
+page that will do it for you.</li>
+</ol>
+<p>Once you’ve been granted commit access, you should be able to check out an LLVM
+tree with an SVN URL of “<a class="reference external" href="https://username@llvm.org/">https://username@llvm.org/</a>...” instead of the normal
+anonymous URL of “<a class="reference external" href="http://llvm.org/">http://llvm.org/</a>...”.  The first time you commit you’ll have
+to type in your password.  Note that you may get a warning from SVN about an
+untrusted key; you can ignore this.  To verify that your commit access works,
+please do a test commit (e.g. change a comment or add a blank line).  Your first
+commit to a repository may require the autogenerated email to be approved by a
+mailing list.  This is normal and will be done when the mailing list owner has
+time.</p>
+<p>If you have recently been granted commit access, these policies apply:</p>
+<ol class="arabic simple">
+<li>You are granted <em>commit-after-approval</em> to all parts of LLVM.  To get
+approval, submit a <a class="reference internal" href="#patch">patch</a> to <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">llvm-commits</a>. When approved,
+you may commit it yourself.</li>
+<li>You are allowed to commit patches without approval which you think are
+obvious. This is clearly a subjective decision — we simply expect you to
+use good judgement.  Examples include: fixing build breakage, reverting
+obviously broken patches, documentation/comment changes, any other minor
+changes.</li>
+<li>You are allowed to commit patches without approval to those portions of LLVM
+that you have contributed or maintain (i.e., have been assigned
+responsibility for), with the proviso that such commits must not break the
+build.  This is a “trust but verify” policy, and commits of this nature are
+reviewed after they are committed.</li>
+<li>Multiple violations of these policies or a single egregious violation may
+cause commit access to be revoked.</li>
+</ol>
+<p>In any case, your changes are still subject to <a class="reference internal" href="#code-review">code review</a> (either before or
+after they are committed, depending on the nature of the change).  You are
+encouraged to review other peoples’ patches as well, but you aren’t required
+to do so.</p>
+</div>
+<div class="section" id="making-a-major-change">
+<span id="discuss-the-change-gather-consensus"></span><h3><a class="toc-backref" href="#id15">Making a Major Change</a><a class="headerlink" href="#making-a-major-change" title="Permalink to this headline">¶</a></h3>
+<p>When a developer begins a major new project with the aim of contributing it back
+to LLVM, they should inform the community with an email to the <a class="reference external" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev</a> email list, to the extent
+possible. The reason for this is to:</p>
+<ol class="arabic simple">
+<li>keep the community informed about future changes to LLVM,</li>
+<li>avoid duplication of effort by preventing multiple parties working on the
+same thing and not knowing about it, and</li>
+<li>ensure that any technical issues around the proposed work are discussed and
+resolved before any significant work is done.</li>
+</ol>
+<p>The design of LLVM is carefully controlled to ensure that all the pieces fit
+together well and are as consistent as possible. If you plan to make a major
+change to the way LLVM works or want to add a major new extension, it is a good
+idea to get consensus with the development community before you start working on
+it.</p>
+<p>Once the design of the new feature is finalized, the work itself should be done
+as a series of <a class="reference internal" href="#incremental-changes">incremental changes</a>, not as a long-term development branch.</p>
+</div>
+<div class="section" id="incremental-development">
+<span id="incremental-changes"></span><h3><a class="toc-backref" href="#id16">Incremental Development</a><a class="headerlink" href="#incremental-development" title="Permalink to this headline">¶</a></h3>
+<p>In the LLVM project, we do all significant changes as a series of incremental
+patches.  We have a strong dislike for huge changes or long-term development
+branches.  Long-term development branches have a number of drawbacks:</p>
+<ol class="arabic simple">
+<li>Branches must have mainline merged into them periodically.  If the branch
+development and mainline development occur in the same pieces of code,
+resolving merge conflicts can take a lot of time.</li>
+<li>Other people in the community tend to ignore work on branches.</li>
+<li>Huge changes (produced when a branch is merged back onto mainline) are
+extremely difficult to <a class="reference internal" href="#code-review">code review</a>.</li>
+<li>Branches are not routinely tested by our nightly tester infrastructure.</li>
+<li>Changes developed as monolithic large changes often don’t work until the
+entire set of changes is done.  Breaking it down into a set of smaller
+changes increases the odds that any of the work will be committed to the main
+repository.</li>
+</ol>
+<p>To address these problems, LLVM uses an incremental development style and we
+require contributors to follow this practice when making a large/invasive
+change.  Some tips:</p>
+<ul class="simple">
+<li>Large/invasive changes usually have a number of secondary changes that are
+required before the big change can be made (e.g. API cleanup, etc).  These
+sorts of changes can often be done before the major change is done,
+independently of that work.</li>
+<li>The remaining inter-related work should be decomposed into unrelated sets of
+changes if possible.  Once this is done, define the first increment and get
+consensus on what the end goal of the change is.</li>
+<li>Each change in the set can be stand alone (e.g. to fix a bug), or part of a
+planned series of changes that works towards the development goal.</li>
+<li>Each change should be kept as small as possible. This simplifies your work
+(into a logical progression), simplifies code review and reduces the chance
+that you will get negative feedback on the change. Small increments also
+facilitate the maintenance of a high quality code base.</li>
+<li>Often, an independent precursor to a big change is to add a new API and slowly
+migrate clients to use the new API.  Each change to use the new API is often
+“obvious” and can be committed without review.  Once the new API is in place
+and used, it is much easier to replace the underlying implementation of the
+API.  This implementation change is logically separate from the API
+change.</li>
+</ul>
+<p>If you are interested in making a large change, and this scares you, please make
+sure to first <a class="reference internal" href="#discuss-the-change-gather-consensus">discuss the change/gather consensus</a> then ask about the best way
+to go about making the change.</p>
+</div>
+<div class="section" id="attribution-of-changes">
+<h3><a class="toc-backref" href="#id17">Attribution of Changes</a><a class="headerlink" href="#attribution-of-changes" title="Permalink to this headline">¶</a></h3>
+<p>When contributors submit a patch to an LLVM project, other developers with
+commit access may commit it for the author once appropriate (based on the
+progression of code review, etc.). When doing so, it is important to retain
+correct attribution of contributions to their contributors. However, we do not
+want the source code to be littered with random attributions “this code written
+by J. Random Hacker” (this is noisy and distracting). In practice, the revision
+control system keeps a perfect history of who changed what, and the CREDITS.txt
+file describes higher-level contributions. If you commit a patch for someone
+else, please say “patch contributed by J. Random Hacker!” in the commit
+message. Overall, please do not add contributor names to the source code.</p>
+<p>Also, don’t commit patches authored by others unless they have submitted the
+patch to the project or you have been authorized to submit them on their behalf
+(you work together and your company authorized you to contribute the patches,
+etc.). The author should first submit them to the relevant project’s commit
+list, development list, or LLVM bug tracker component. If someone sends you
+a patch privately, encourage them to submit it to the appropriate list first.</p>
+</div>
+</div>
+<div class="section" id="copyright-license-and-patents">
+<span id="copyright-license-patents"></span><h2><a class="toc-backref" href="#id18">Copyright, License, and Patents</a><a class="headerlink" href="#copyright-license-and-patents" title="Permalink to this headline">¶</a></h2>
+<div class="admonition note">
+<p class="first admonition-title">Note</p>
+<p class="last">This section deals with legal matters but does not provide legal advice.  We
+are not lawyers — please seek legal counsel from an attorney.</p>
+</div>
+<p>This section addresses the issues of copyright, license and patents for the LLVM
+project.  The copyright for the code is held by the individual contributors of
+the code and the terms of its license to LLVM users and developers is the
+<a class="reference external" href="http://www.opensource.org/licenses/UoI-NCSA.php">University of Illinois/NCSA Open Source License</a> (with portions dual licensed
+under the <a class="reference external" href="http://www.opensource.org/licenses/mit-license.php">MIT License</a>,
+see below).  As contributor to the LLVM project, you agree to allow any
+contributions to the project to licensed under these terms.</p>
+<div class="section" id="copyright">
+<h3><a class="toc-backref" href="#id19">Copyright</a><a class="headerlink" href="#copyright" title="Permalink to this headline">¶</a></h3>
+<p>The LLVM project does not require copyright assignments, which means that the
+copyright for the code in the project is held by its respective contributors who
+have each agreed to release their contributed code under the terms of the <a class="reference internal" href="#llvm-license">LLVM
+License</a>.</p>
+<p>An implication of this is that the LLVM license is unlikely to ever change:
+changing it would require tracking down all the contributors to LLVM and getting
+them to agree that a license change is acceptable for their contribution.  Since
+there are no plans to change the license, this is not a cause for concern.</p>
+<p>As a contributor to the project, this means that you (or your company) retain
+ownership of the code you contribute, that it cannot be used in a way that
+contradicts the license (which is a liberal BSD-style license), and that the
+license for your contributions won’t change without your approval in the
+future.</p>
+</div>
+<div class="section" id="license">
+<span id="llvm-license"></span><h3><a class="toc-backref" href="#id20">License</a><a class="headerlink" href="#license" title="Permalink to this headline">¶</a></h3>
+<p>We intend to keep LLVM perpetually open source and to use a liberal open source
+license. <strong>As a contributor to the project, you agree that any contributions be
+licensed under the terms of the corresponding subproject.</strong> All of the code in
+LLVM is available under the <a class="reference external" href="http://www.opensource.org/licenses/UoI-NCSA.php">University of Illinois/NCSA Open Source License</a>, which boils down to
+this:</p>
+<ul class="simple">
+<li>You can freely distribute LLVM.</li>
+<li>You must retain the copyright notice if you redistribute LLVM.</li>
+<li>Binaries derived from LLVM must reproduce the copyright notice (e.g. in an
+included readme file).</li>
+<li>You can’t use our names to promote your LLVM derived products.</li>
+<li>There’s no warranty on LLVM at all.</li>
+</ul>
+<p>We believe this fosters the widest adoption of LLVM because it <strong>allows
+commercial products to be derived from LLVM</strong> with few restrictions and without
+a requirement for making any derived works also open source (i.e.  LLVM’s
+license is not a “copyleft” license like the GPL). We suggest that you read the
+<a class="reference external" href="http://www.opensource.org/licenses/UoI-NCSA.php">License</a> if further
+clarification is needed.</p>
+<p>In addition to the UIUC license, the runtime library components of LLVM
+(<strong>compiler_rt, libc++, and libclc</strong>) are also licensed under the <a class="reference external" href="http://www.opensource.org/licenses/mit-license.php">MIT License</a>, which does not contain
+the binary redistribution clause.  As a user of these runtime libraries, it
+means that you can choose to use the code under either license (and thus don’t
+need the binary redistribution clause), and as a contributor to the code that
+you agree that any contributions to these libraries be licensed under both
+licenses.  We feel that this is important for runtime libraries, because they
+are implicitly linked into applications and therefore should not subject those
+applications to the binary redistribution clause. This also means that it is ok
+to move code from (e.g.)  libc++ to the LLVM core without concern, but that code
+cannot be moved from the LLVM core to libc++ without the copyright owner’s
+permission.</p>
+<p>Note that the LLVM Project does distribute dragonegg, <strong>which is
+GPL.</strong> This means that anything “linked” into dragonegg must itself be compatible
+with the GPL, and must be releasable under the terms of the GPL.  This implies
+that <strong>any code linked into dragonegg and distributed to others may be subject to
+the viral aspects of the GPL</strong> (for example, a proprietary code generator linked
+into dragonegg must be made available under the GPL).  This is not a problem for
+code already distributed under a more liberal license (like the UIUC license),
+and GPL-containing subprojects are kept in separate SVN repositories whose
+LICENSE.txt files specifically indicate that they contain GPL code.</p>
+<p>We have no plans to change the license of LLVM.  If you have questions or
+comments about the license, please contact the <a class="reference external" href="mailto:llvmdev%40cs.uiuc.edu">LLVM Developer's Mailing
+List</a>.</p>
+</div>
+<div class="section" id="patents">
+<h3><a class="toc-backref" href="#id21">Patents</a><a class="headerlink" href="#patents" title="Permalink to this headline">¶</a></h3>
+<p>To the best of our knowledge, LLVM does not infringe on any patents (we have
+actually removed code from LLVM in the past that was found to infringe).  Having
+code in LLVM that infringes on patents would violate an important goal of the
+project by making it hard or impossible to reuse the code for arbitrary purposes
+(including commercial use).</p>
+<p>When contributing code, we expect contributors to notify us of any potential for
+patent-related trouble with their changes (including from third parties).  If
+you or your employer own the rights to a patent and would like to contribute
+code to LLVM that relies on it, we require that the copyright owner sign an
+agreement that allows any other user of LLVM to freely use your patent.  Please
+contact the <a class="reference external" href="mailto:llvm-oversight%40cs.uiuc.edu">oversight group</a> for more
+details.</p>
+</div>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="MakefileGuide.html" title="LLVM Makefile Guide"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="BigEndianNEON.html" title="Using ARM NEON instructions in big endian mode"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/ExceptionHandling.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/ExceptionHandling.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/ExceptionHandling.html (added)
+++ www-releases/trunk/3.5.1/docs/ExceptionHandling.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,493 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>Exception Handling in LLVM — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="LLVM Link Time Optimization: Design and Implementation" href="LinkTimeOptimization.html" />
+    <link rel="prev" title="The LLVM Target-Independent Code Generator" href="CodeGenerator.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">
+      <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="LinkTimeOptimization.html" title="LLVM Link Time Optimization: Design and Implementation"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="CodeGenerator.html" title="The LLVM Target-Independent Code Generator"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="exception-handling-in-llvm">
+<h1>Exception Handling in LLVM<a class="headerlink" href="#exception-handling-in-llvm" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#introduction" id="id5">Introduction</a><ul>
+<li><a class="reference internal" href="#itanium-abi-zero-cost-exception-handling" id="id6">Itanium ABI Zero-cost Exception Handling</a></li>
+<li><a class="reference internal" href="#setjmp-longjmp-exception-handling" id="id7">Setjmp/Longjmp Exception Handling</a></li>
+<li><a class="reference internal" href="#overview" id="id8">Overview</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#llvm-code-generation" id="id9">LLVM Code Generation</a><ul>
+<li><a class="reference internal" href="#throw" id="id10">Throw</a></li>
+<li><a class="reference internal" href="#try-catch" id="id11">Try/Catch</a></li>
+<li><a class="reference internal" href="#cleanups" id="id12">Cleanups</a></li>
+<li><a class="reference internal" href="#throw-filters" id="id13">Throw Filters</a></li>
+<li><a class="reference internal" href="#restrictions" id="id14">Restrictions</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#exception-handling-intrinsics" id="id15">Exception Handling Intrinsics</a><ul>
+<li><a class="reference internal" href="#llvm-eh-typeid-for" id="id16"><tt class="docutils literal"><span class="pre">llvm.eh.typeid.for</span></tt></a></li>
+<li><a class="reference internal" href="#sjlj-intrinsics" id="id17">SJLJ Intrinsics</a><ul>
+<li><a class="reference internal" href="#llvm-eh-sjlj-setjmp" id="id18"><tt class="docutils literal"><span class="pre">llvm.eh.sjlj.setjmp</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-eh-sjlj-longjmp" id="id19"><tt class="docutils literal"><span class="pre">llvm.eh.sjlj.longjmp</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-eh-sjlj-lsda" id="id20"><tt class="docutils literal"><span class="pre">llvm.eh.sjlj.lsda</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-eh-sjlj-callsite" id="id21"><tt class="docutils literal"><span class="pre">llvm.eh.sjlj.callsite</span></tt></a></li>
+</ul>
+</li>
+</ul>
+</li>
+<li><a class="reference internal" href="#asm-table-formats" id="id22">Asm Table Formats</a><ul>
+<li><a class="reference internal" href="#exception-handling-frame" id="id23">Exception Handling Frame</a></li>
+<li><a class="reference internal" href="#exception-tables" id="id24">Exception Tables</a></li>
+</ul>
+</li>
+</ul>
+</div>
+<div class="section" id="introduction">
+<h2><a class="toc-backref" href="#id5">Introduction</a><a class="headerlink" href="#introduction" title="Permalink to this headline">¶</a></h2>
+<p>This document is the central repository for all information pertaining to
+exception handling in LLVM.  It describes the format that LLVM exception
+handling information takes, which is useful for those interested in creating
+front-ends or dealing directly with the information.  Further, this document
+provides specific examples of what exception handling information is used for in
+C and C++.</p>
+<div class="section" id="itanium-abi-zero-cost-exception-handling">
+<h3><a class="toc-backref" href="#id6">Itanium ABI Zero-cost Exception Handling</a><a class="headerlink" href="#itanium-abi-zero-cost-exception-handling" title="Permalink to this headline">¶</a></h3>
+<p>Exception handling for most programming languages is designed to recover from
+conditions that rarely occur during general use of an application.  To that end,
+exception handling should not interfere with the main flow of an application’s
+algorithm by performing checkpointing tasks, such as saving the current pc or
+register state.</p>
+<p>The Itanium ABI Exception Handling Specification defines a methodology for
+providing outlying data in the form of exception tables without inlining
+speculative exception handling code in the flow of an application’s main
+algorithm.  Thus, the specification is said to add “zero-cost” to the normal
+execution of an application.</p>
+<p>A more complete description of the Itanium ABI exception handling runtime
+support of can be found at <a class="reference external" href="http://mentorembedded.github.com/cxx-abi/abi-eh.html">Itanium C++ ABI: Exception Handling</a>. A description of the
+exception frame format can be found at <a class="reference external" href="http://refspecs.linuxfoundation.org/LSB_3.0.0/LSB-Core-generic/LSB-Core-generic/ehframechpt.html">Exception Frames</a>,
+with details of the DWARF 4 specification at <a class="reference external" href="http://dwarfstd.org/Dwarf4Std.php">DWARF 4 Standard</a>.  A description for the C++ exception
+table formats can be found at <a class="reference external" href="http://mentorembedded.github.com/cxx-abi/exceptions.pdf">Exception Handling Tables</a>.</p>
+</div>
+<div class="section" id="setjmp-longjmp-exception-handling">
+<h3><a class="toc-backref" href="#id7">Setjmp/Longjmp Exception Handling</a><a class="headerlink" href="#setjmp-longjmp-exception-handling" title="Permalink to this headline">¶</a></h3>
+<p>Setjmp/Longjmp (SJLJ) based exception handling uses LLVM intrinsics
+<a class="reference internal" href="#llvm-eh-sjlj-setjmp">llvm.eh.sjlj.setjmp</a> and <a class="reference internal" href="#llvm-eh-sjlj-longjmp">llvm.eh.sjlj.longjmp</a> to handle control flow for
+exception handling.</p>
+<p>For each function which does exception processing — be it <tt class="docutils literal"><span class="pre">try</span></tt>/<tt class="docutils literal"><span class="pre">catch</span></tt>
+blocks or cleanups — that function registers itself on a global frame
+list. When exceptions are unwinding, the runtime uses this list to identify
+which functions need processing.</p>
+<p>Landing pad selection is encoded in the call site entry of the function
+context. The runtime returns to the function via <a class="reference internal" href="#llvm-eh-sjlj-longjmp">llvm.eh.sjlj.longjmp</a>, where
+a switch table transfers control to the appropriate landing pad based on the
+index stored in the function context.</p>
+<p>In contrast to DWARF exception handling, which encodes exception regions and
+frame information in out-of-line tables, SJLJ exception handling builds and
+removes the unwind frame context at runtime. This results in faster exception
+handling at the expense of slower execution when no exceptions are thrown. As
+exceptions are, by their nature, intended for uncommon code paths, DWARF
+exception handling is generally preferred to SJLJ.</p>
+</div>
+<div class="section" id="overview">
+<h3><a class="toc-backref" href="#id8">Overview</a><a class="headerlink" href="#overview" title="Permalink to this headline">¶</a></h3>
+<p>When an exception is thrown in LLVM code, the runtime does its best to find a
+handler suited to processing the circumstance.</p>
+<p>The runtime first attempts to find an <em>exception frame</em> corresponding to the
+function where the exception was thrown.  If the programming language supports
+exception handling (e.g. C++), the exception frame contains a reference to an
+exception table describing how to process the exception.  If the language does
+not support exception handling (e.g. C), or if the exception needs to be
+forwarded to a prior activation, the exception frame contains information about
+how to unwind the current activation and restore the state of the prior
+activation.  This process is repeated until the exception is handled. If the
+exception is not handled and no activations remain, then the application is
+terminated with an appropriate error message.</p>
+<p>Because different programming languages have different behaviors when handling
+exceptions, the exception handling ABI provides a mechanism for
+supplying <em>personalities</em>. An exception handling personality is defined by
+way of a <em>personality function</em> (e.g. <tt class="docutils literal"><span class="pre">__gxx_personality_v0</span></tt> in C++),
+which receives the context of the exception, an <em>exception structure</em>
+containing the exception object type and value, and a reference to the exception
+table for the current function.  The personality function for the current
+compile unit is specified in a <em>common exception frame</em>.</p>
+<p>The organization of an exception table is language dependent. For C++, an
+exception table is organized as a series of code ranges defining what to do if
+an exception occurs in that range. Typically, the information associated with a
+range defines which types of exception objects (using C++ <em>type info</em>) that are
+handled in that range, and an associated action that should take place. Actions
+typically pass control to a <em>landing pad</em>.</p>
+<p>A landing pad corresponds roughly to the code found in the <tt class="docutils literal"><span class="pre">catch</span></tt> portion of
+a <tt class="docutils literal"><span class="pre">try</span></tt>/<tt class="docutils literal"><span class="pre">catch</span></tt> sequence. When execution resumes at a landing pad, it
+receives an <em>exception structure</em> and a <em>selector value</em> corresponding to the
+<em>type</em> of exception thrown. The selector is then used to determine which <em>catch</em>
+should actually process the exception.</p>
+</div>
+</div>
+<div class="section" id="llvm-code-generation">
+<h2><a class="toc-backref" href="#id9">LLVM Code Generation</a><a class="headerlink" href="#llvm-code-generation" title="Permalink to this headline">¶</a></h2>
+<p>From a C++ developer’s perspective, exceptions are defined in terms of the
+<tt class="docutils literal"><span class="pre">throw</span></tt> and <tt class="docutils literal"><span class="pre">try</span></tt>/<tt class="docutils literal"><span class="pre">catch</span></tt> statements. In this section we will describe the
+implementation of LLVM exception handling in terms of C++ examples.</p>
+<div class="section" id="throw">
+<h3><a class="toc-backref" href="#id10">Throw</a><a class="headerlink" href="#throw" title="Permalink to this headline">¶</a></h3>
+<p>Languages that support exception handling typically provide a <tt class="docutils literal"><span class="pre">throw</span></tt>
+operation to initiate the exception process. Internally, a <tt class="docutils literal"><span class="pre">throw</span></tt> operation
+breaks down into two steps.</p>
+<ol class="arabic simple">
+<li>A request is made to allocate exception space for an exception structure.
+This structure needs to survive beyond the current activation. This structure
+will contain the type and value of the object being thrown.</li>
+<li>A call is made to the runtime to raise the exception, passing the exception
+structure as an argument.</li>
+</ol>
+<p>In C++, the allocation of the exception structure is done by the
+<tt class="docutils literal"><span class="pre">__cxa_allocate_exception</span></tt> runtime function. The exception raising is handled
+by <tt class="docutils literal"><span class="pre">__cxa_throw</span></tt>. The type of the exception is represented using a C++ RTTI
+structure.</p>
+</div>
+<div class="section" id="try-catch">
+<h3><a class="toc-backref" href="#id11">Try/Catch</a><a class="headerlink" href="#try-catch" title="Permalink to this headline">¶</a></h3>
+<p>A call within the scope of a <em>try</em> statement can potentially raise an
+exception. In those circumstances, the LLVM C++ front-end replaces the call with
+an <tt class="docutils literal"><span class="pre">invoke</span></tt> instruction. Unlike a call, the <tt class="docutils literal"><span class="pre">invoke</span></tt> has two potential
+continuation points:</p>
+<ol class="arabic simple">
+<li>where to continue when the call succeeds as per normal, and</li>
+<li>where to continue if the call raises an exception, either by a throw or the
+unwinding of a throw</li>
+</ol>
+<p>The term used to define the place where an <tt class="docutils literal"><span class="pre">invoke</span></tt> continues after an
+exception is called a <em>landing pad</em>. LLVM landing pads are conceptually
+alternative function entry points where an exception structure reference and a
+type info index are passed in as arguments. The landing pad saves the exception
+structure reference and then proceeds to select the catch block that corresponds
+to the type info of the exception object.</p>
+<p>The LLVM <a class="reference internal" href="LangRef.html#i-landingpad"><em>‘landingpad’ Instruction</em></a> is used to convey information about the landing
+pad to the back end. For C++, the <tt class="docutils literal"><span class="pre">landingpad</span></tt> instruction returns a pointer
+and integer pair corresponding to the pointer to the <em>exception structure</em> and
+the <em>selector value</em> respectively.</p>
+<p>The <tt class="docutils literal"><span class="pre">landingpad</span></tt> instruction takes a reference to the personality function to
+be used for this <tt class="docutils literal"><span class="pre">try</span></tt>/<tt class="docutils literal"><span class="pre">catch</span></tt> sequence. The remainder of the instruction is
+a list of <em>cleanup</em>, <em>catch</em>, and <em>filter</em> clauses. The exception is tested
+against the clauses sequentially from first to last. The clauses have the
+following meanings:</p>
+<ul>
+<li><p class="first"><tt class="docutils literal"><span class="pre">catch</span> <span class="pre"><type></span> <span class="pre">@ExcType</span></tt></p>
+<ul class="simple">
+<li>This clause means that the landingpad block should be entered if the
+exception being thrown is of type <tt class="docutils literal"><span class="pre">@ExcType</span></tt> or a subtype of
+<tt class="docutils literal"><span class="pre">@ExcType</span></tt>. For C++, <tt class="docutils literal"><span class="pre">@ExcType</span></tt> is a pointer to the <tt class="docutils literal"><span class="pre">std::type_info</span></tt>
+object (an RTTI object) representing the C++ exception type.</li>
+<li>If <tt class="docutils literal"><span class="pre">@ExcType</span></tt> is <tt class="docutils literal"><span class="pre">null</span></tt>, any exception matches, so the landingpad
+should always be entered. This is used for C++ catch-all blocks (“<tt class="docutils literal"><span class="pre">catch</span>
+<span class="pre">(...)</span></tt>”).</li>
+<li>When this clause is matched, the selector value will be equal to the value
+returned by “<tt class="docutils literal"><span class="pre">@llvm.eh.typeid.for(i8*</span> <span class="pre">@ExcType)</span></tt>”. This will always be a
+positive value.</li>
+</ul>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">filter</span> <span class="pre"><type></span> <span class="pre">[<type></span> <span class="pre">@ExcType1,</span> <span class="pre">...,</span> <span class="pre"><type></span> <span class="pre">@ExcTypeN]</span></tt></p>
+<ul class="simple">
+<li>This clause means that the landingpad should be entered if the exception
+being thrown does <em>not</em> match any of the types in the list (which, for C++,
+are again specified as <tt class="docutils literal"><span class="pre">std::type_info</span></tt> pointers).</li>
+<li>C++ front-ends use this to implement C++ exception specifications, such as
+“<tt class="docutils literal"><span class="pre">void</span> <span class="pre">foo()</span> <span class="pre">throw</span> <span class="pre">(ExcType1,</span> <span class="pre">...,</span> <span class="pre">ExcTypeN)</span> <span class="pre">{</span> <span class="pre">...</span> <span class="pre">}</span></tt>”.</li>
+<li>When this clause is matched, the selector value will be negative.</li>
+<li>The array argument to <tt class="docutils literal"><span class="pre">filter</span></tt> may be empty; for example, “<tt class="docutils literal"><span class="pre">[0</span> <span class="pre">x</span> <span class="pre">i8**]</span>
+<span class="pre">undef</span></tt>”. This means that the landingpad should always be entered. (Note
+that such a <tt class="docutils literal"><span class="pre">filter</span></tt> would not be equivalent to “<tt class="docutils literal"><span class="pre">catch</span> <span class="pre">i8*</span> <span class="pre">null</span></tt>”,
+because <tt class="docutils literal"><span class="pre">filter</span></tt> and <tt class="docutils literal"><span class="pre">catch</span></tt> produce negative and positive selector
+values respectively.)</li>
+</ul>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">cleanup</span></tt></p>
+<ul>
+<li><p class="first">This clause means that the landingpad should always be entered.</p>
+</li>
+<li><p class="first">C++ front-ends use this for calling objects’ destructors.</p>
+</li>
+<li><p class="first">When this clause is matched, the selector value will be zero.</p>
+</li>
+<li><p class="first">The runtime may treat “<tt class="docutils literal"><span class="pre">cleanup</span></tt>” differently from “<tt class="docutils literal"><span class="pre">catch</span> <span class="pre"><type></span>
+<span class="pre">null</span></tt>”.</p>
+<p>In C++, if an unhandled exception occurs, the language runtime will call
+<tt class="docutils literal"><span class="pre">std::terminate()</span></tt>, but it is implementation-defined whether the runtime
+unwinds the stack and calls object destructors first. For example, the GNU
+C++ unwinder does not call object destructors when an unhandled exception
+occurs. The reason for this is to improve debuggability: it ensures that
+<tt class="docutils literal"><span class="pre">std::terminate()</span></tt> is called from the context of the <tt class="docutils literal"><span class="pre">throw</span></tt>, so that
+this context is not lost by unwinding the stack. A runtime will typically
+implement this by searching for a matching non-<tt class="docutils literal"><span class="pre">cleanup</span></tt> clause, and
+aborting if it does not find one, before entering any landingpad blocks.</p>
+</li>
+</ul>
+</li>
+</ul>
+<p>Once the landing pad has the type info selector, the code branches to the code
+for the first catch. The catch then checks the value of the type info selector
+against the index of type info for that catch.  Since the type info index is not
+known until all the type infos have been gathered in the backend, the catch code
+must call the <a class="reference internal" href="#llvm-eh-typeid-for">llvm.eh.typeid.for</a> intrinsic to determine the index for a given
+type info. If the catch fails to match the selector then control is passed on to
+the next catch.</p>
+<p>Finally, the entry and exit of catch code is bracketed with calls to
+<tt class="docutils literal"><span class="pre">__cxa_begin_catch</span></tt> and <tt class="docutils literal"><span class="pre">__cxa_end_catch</span></tt>.</p>
+<ul>
+<li><p class="first"><tt class="docutils literal"><span class="pre">__cxa_begin_catch</span></tt> takes an exception structure reference as an argument
+and returns the value of the exception object.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">__cxa_end_catch</span></tt> takes no arguments. This function:</p>
+<ol class="arabic simple">
+<li>Locates the most recently caught exception and decrements its handler
+count,</li>
+<li>Removes the exception from the <em>caught</em> stack if the handler count goes to
+zero, and</li>
+<li>Destroys the exception if the handler count goes to zero and the exception
+was not re-thrown by throw.</li>
+</ol>
+<div class="admonition note">
+<p class="first admonition-title">Note</p>
+<p class="last">a rethrow from within the catch may replace this call with a
+<tt class="docutils literal"><span class="pre">__cxa_rethrow</span></tt>.</p>
+</div>
+</li>
+</ul>
+</div>
+<div class="section" id="cleanups">
+<h3><a class="toc-backref" href="#id12">Cleanups</a><a class="headerlink" href="#cleanups" title="Permalink to this headline">¶</a></h3>
+<p>A cleanup is extra code which needs to be run as part of unwinding a scope.  C++
+destructors are a typical example, but other languages and language extensions
+provide a variety of different kinds of cleanups. In general, a landing pad may
+need to run arbitrary amounts of cleanup code before actually entering a catch
+block. To indicate the presence of cleanups, a <a class="reference internal" href="LangRef.html#i-landingpad"><em>‘landingpad’ Instruction</em></a> should have
+a <em>cleanup</em> clause.  Otherwise, the unwinder will not stop at the landing pad if
+there are no catches or filters that require it to.</p>
+<div class="admonition note">
+<p class="first admonition-title">Note</p>
+<p class="last">Do not allow a new exception to propagate out of the execution of a
+cleanup. This can corrupt the internal state of the unwinder.  Different
+languages describe different high-level semantics for these situations: for
+example, C++ requires that the process be terminated, whereas Ada cancels both
+exceptions and throws a third.</p>
+</div>
+<p>When all cleanups are finished, if the exception is not handled by the current
+function, resume unwinding by calling the <a class="reference external" href="LangRef.html#i_resume">resume
+instruction</a>, passing in the result of the
+<tt class="docutils literal"><span class="pre">landingpad</span></tt> instruction for the original landing pad.</p>
+</div>
+<div class="section" id="throw-filters">
+<h3><a class="toc-backref" href="#id13">Throw Filters</a><a class="headerlink" href="#throw-filters" title="Permalink to this headline">¶</a></h3>
+<p>C++ allows the specification of which exception types may be thrown from a
+function. To represent this, a top level landing pad may exist to filter out
+invalid types. To express this in LLVM code the <a class="reference internal" href="LangRef.html#i-landingpad"><em>‘landingpad’ Instruction</em></a> will have a
+filter clause. The clause consists of an array of type infos.
+<tt class="docutils literal"><span class="pre">landingpad</span></tt> will return a negative value
+if the exception does not match any of the type infos. If no match is found then
+a call to <tt class="docutils literal"><span class="pre">__cxa_call_unexpected</span></tt> should be made, otherwise
+<tt class="docutils literal"><span class="pre">_Unwind_Resume</span></tt>.  Each of these functions requires a reference to the
+exception structure.  Note that the most general form of a <tt class="docutils literal"><span class="pre">landingpad</span></tt>
+instruction can have any number of catch, cleanup, and filter clauses (though
+having more than one cleanup is pointless). The LLVM C++ front-end can generate
+such <tt class="docutils literal"><span class="pre">landingpad</span></tt> instructions due to inlining creating nested exception
+handling scopes.</p>
+</div>
+<div class="section" id="restrictions">
+<span id="undefined"></span><h3><a class="toc-backref" href="#id14">Restrictions</a><a class="headerlink" href="#restrictions" title="Permalink to this headline">¶</a></h3>
+<p>The unwinder delegates the decision of whether to stop in a call frame to that
+call frame’s language-specific personality function. Not all unwinders guarantee
+that they will stop to perform cleanups. For example, the GNU C++ unwinder
+doesn’t do so unless the exception is actually caught somewhere further up the
+stack.</p>
+<p>In order for inlining to behave correctly, landing pads must be prepared to
+handle selector results that they did not originally advertise. Suppose that a
+function catches exceptions of type <tt class="docutils literal"><span class="pre">A</span></tt>, and it’s inlined into a function that
+catches exceptions of type <tt class="docutils literal"><span class="pre">B</span></tt>. The inliner will update the <tt class="docutils literal"><span class="pre">landingpad</span></tt>
+instruction for the inlined landing pad to include the fact that <tt class="docutils literal"><span class="pre">B</span></tt> is also
+caught. If that landing pad assumes that it will only be entered to catch an
+<tt class="docutils literal"><span class="pre">A</span></tt>, it’s in for a rude awakening.  Consequently, landing pads must test for
+the selector results they understand and then resume exception propagation with
+the <a class="reference external" href="LangRef.html#i_resume">resume instruction</a> if none of the conditions
+match.</p>
+</div>
+</div>
+<div class="section" id="exception-handling-intrinsics">
+<h2><a class="toc-backref" href="#id15">Exception Handling Intrinsics</a><a class="headerlink" href="#exception-handling-intrinsics" title="Permalink to this headline">¶</a></h2>
+<p>In addition to the <tt class="docutils literal"><span class="pre">landingpad</span></tt> and <tt class="docutils literal"><span class="pre">resume</span></tt> instructions, LLVM uses several
+intrinsic functions (name prefixed with <tt class="docutils literal"><span class="pre">llvm.eh</span></tt>) to provide exception
+handling information at various points in generated code.</p>
+<div class="section" id="llvm-eh-typeid-for">
+<span id="id2"></span><h3><a class="toc-backref" href="#id16"><tt class="docutils literal"><span class="pre">llvm.eh.typeid.for</span></tt></a><a class="headerlink" href="#llvm-eh-typeid-for" title="Permalink to this headline">¶</a></h3>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">i32</span> <span class="vg">@llvm.eh.typeid.for</span><span class="p">(</span><span class="k">i8</span><span class="p">*</span> <span class="nv">%type_info</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>This intrinsic returns the type info index in the exception table of the current
+function.  This value can be used to compare against the result of
+<tt class="docutils literal"><span class="pre">landingpad</span></tt> instruction.  The single argument is a reference to a type info.</p>
+<p>Uses of this intrinsic are generated by the C++ front-end.</p>
+</div>
+<div class="section" id="sjlj-intrinsics">
+<h3><a class="toc-backref" href="#id17">SJLJ Intrinsics</a><a class="headerlink" href="#sjlj-intrinsics" title="Permalink to this headline">¶</a></h3>
+<p>The <tt class="docutils literal"><span class="pre">llvm.eh.sjlj</span></tt> intrinsics are used internally within LLVM’s
+backend.  Uses of them are generated by the backend’s
+<tt class="docutils literal"><span class="pre">SjLjEHPrepare</span></tt> pass.</p>
+<div class="section" id="llvm-eh-sjlj-setjmp">
+<span id="id3"></span><h4><a class="toc-backref" href="#id18"><tt class="docutils literal"><span class="pre">llvm.eh.sjlj.setjmp</span></tt></a><a class="headerlink" href="#llvm-eh-sjlj-setjmp" title="Permalink to this headline">¶</a></h4>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">i32</span> <span class="vg">@llvm.eh.sjlj.setjmp</span><span class="p">(</span><span class="k">i8</span><span class="p">*</span> <span class="nv">%setjmp_buf</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>For SJLJ based exception handling, this intrinsic forces register saving for the
+current function and stores the address of the following instruction for use as
+a destination address by <a class="reference internal" href="#llvm-eh-sjlj-longjmp">llvm.eh.sjlj.longjmp</a>. The buffer format and the
+overall functioning of this intrinsic is compatible with the GCC
+<tt class="docutils literal"><span class="pre">__builtin_setjmp</span></tt> implementation allowing code built with the clang and GCC
+to interoperate.</p>
+<p>The single parameter is a pointer to a five word buffer in which the calling
+context is saved. The front end places the frame pointer in the first word, and
+the target implementation of this intrinsic should place the destination address
+for a <a class="reference internal" href="#llvm-eh-sjlj-longjmp">llvm.eh.sjlj.longjmp</a> in the second word. The following three words are
+available for use in a target-specific manner.</p>
+</div>
+<div class="section" id="llvm-eh-sjlj-longjmp">
+<span id="id4"></span><h4><a class="toc-backref" href="#id19"><tt class="docutils literal"><span class="pre">llvm.eh.sjlj.longjmp</span></tt></a><a class="headerlink" href="#llvm-eh-sjlj-longjmp" title="Permalink to this headline">¶</a></h4>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="kt">void</span> <span class="vg">@llvm.eh.sjlj.longjmp</span><span class="p">(</span><span class="k">i8</span><span class="p">*</span> <span class="nv">%setjmp_buf</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>For SJLJ based exception handling, the <tt class="docutils literal"><span class="pre">llvm.eh.sjlj.longjmp</span></tt> intrinsic is
+used to implement <tt class="docutils literal"><span class="pre">__builtin_longjmp()</span></tt>. The single parameter is a pointer to
+a buffer populated by <a class="reference internal" href="#llvm-eh-sjlj-setjmp">llvm.eh.sjlj.setjmp</a>. The frame pointer and stack
+pointer are restored from the buffer, then control is transferred to the
+destination address.</p>
+</div>
+<div class="section" id="llvm-eh-sjlj-lsda">
+<h4><a class="toc-backref" href="#id20"><tt class="docutils literal"><span class="pre">llvm.eh.sjlj.lsda</span></tt></a><a class="headerlink" href="#llvm-eh-sjlj-lsda" title="Permalink to this headline">¶</a></h4>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">i8</span><span class="p">*</span> <span class="vg">@llvm.eh.sjlj.lsda</span><span class="p">()</span>
+</pre></div>
+</div>
+<p>For SJLJ based exception handling, the <tt class="docutils literal"><span class="pre">llvm.eh.sjlj.lsda</span></tt> intrinsic returns
+the address of the Language Specific Data Area (LSDA) for the current
+function. The SJLJ front-end code stores this address in the exception handling
+function context for use by the runtime.</p>
+</div>
+<div class="section" id="llvm-eh-sjlj-callsite">
+<h4><a class="toc-backref" href="#id21"><tt class="docutils literal"><span class="pre">llvm.eh.sjlj.callsite</span></tt></a><a class="headerlink" href="#llvm-eh-sjlj-callsite" title="Permalink to this headline">¶</a></h4>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="kt">void</span> <span class="vg">@llvm.eh.sjlj.callsite</span><span class="p">(</span><span class="k">i32</span> <span class="nv">%call_site_num</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>For SJLJ based exception handling, the <tt class="docutils literal"><span class="pre">llvm.eh.sjlj.callsite</span></tt> intrinsic
+identifies the callsite value associated with the following <tt class="docutils literal"><span class="pre">invoke</span></tt>
+instruction. This is used to ensure that landing pad entries in the LSDA are
+generated in matching order.</p>
+</div>
+</div>
+</div>
+<div class="section" id="asm-table-formats">
+<h2><a class="toc-backref" href="#id22">Asm Table Formats</a><a class="headerlink" href="#asm-table-formats" title="Permalink to this headline">¶</a></h2>
+<p>There are two tables that are used by the exception handling runtime to
+determine which actions should be taken when an exception is thrown.</p>
+<div class="section" id="exception-handling-frame">
+<h3><a class="toc-backref" href="#id23">Exception Handling Frame</a><a class="headerlink" href="#exception-handling-frame" title="Permalink to this headline">¶</a></h3>
+<p>An exception handling frame <tt class="docutils literal"><span class="pre">eh_frame</span></tt> is very similar to the unwind frame
+used by DWARF debug info. The frame contains all the information necessary to
+tear down the current frame and restore the state of the prior frame. There is
+an exception handling frame for each function in a compile unit, plus a common
+exception handling frame that defines information common to all functions in the
+unit.</p>
+</div>
+<div class="section" id="exception-tables">
+<h3><a class="toc-backref" href="#id24">Exception Tables</a><a class="headerlink" href="#exception-tables" title="Permalink to this headline">¶</a></h3>
+<p>An exception table contains information about what actions to take when an
+exception is thrown in a particular part of a function’s code. There is one
+exception table per function, except leaf functions and functions that have
+calls only to non-throwing functions. They do not need an exception table.</p>
+</div>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="LinkTimeOptimization.html" title="LLVM Link Time Optimization: Design and Implementation"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="CodeGenerator.html" title="The LLVM Target-Independent Code Generator"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/ExtendingLLVM.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/ExtendingLLVM.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/ExtendingLLVM.html (added)
+++ www-releases/trunk/3.5.1/docs/ExtendingLLVM.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,363 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>Extending LLVM: Adding instructions, intrinsics, types, etc. — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="How to set up LLVM-style RTTI for your class hierarchy" href="HowToSetUpLLVMStyleRTTI.html" />
+    <link rel="prev" title="Architecture & Platform Information for Compiler Writers" href="CompilerWriterInfo.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">
+      <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="HowToSetUpLLVMStyleRTTI.html" title="How to set up LLVM-style RTTI for your class hierarchy"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="CompilerWriterInfo.html" title="Architecture & Platform Information for Compiler Writers"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="extending-llvm-adding-instructions-intrinsics-types-etc">
+<h1>Extending LLVM: Adding instructions, intrinsics, types, etc.<a class="headerlink" href="#extending-llvm-adding-instructions-intrinsics-types-etc" title="Permalink to this headline">¶</a></h1>
+<div class="section" id="introduction-and-warning">
+<h2>Introduction and Warning<a class="headerlink" href="#introduction-and-warning" title="Permalink to this headline">¶</a></h2>
+<p>During the course of using LLVM, you may wish to customize it for your research
+project or for experimentation. At this point, you may realize that you need to
+add something to LLVM, whether it be a new fundamental type, a new intrinsic
+function, or a whole new instruction.</p>
+<p>When you come to this realization, stop and think. Do you really need to extend
+LLVM? Is it a new fundamental capability that LLVM does not support at its
+current incarnation or can it be synthesized from already pre-existing LLVM
+elements? If you are not sure, ask on the <a class="reference external" href="http://mail.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVM-dev</a> list. The reason is that
+extending LLVM will get involved as you need to update all the different passes
+that you intend to use with your extension, and there are <tt class="docutils literal"><span class="pre">many</span></tt> LLVM analyses
+and transformations, so it may be quite a bit of work.</p>
+<p>Adding an <a class="reference internal" href="#intrinsic-function">intrinsic function</a> is far easier than adding an
+instruction, and is transparent to optimization passes.  If your added
+functionality can be expressed as a function call, an intrinsic function is the
+method of choice for LLVM extension.</p>
+<p>Before you invest a significant amount of effort into a non-trivial extension,
+<strong>ask on the list</strong> if what you are looking to do can be done with
+already-existing infrastructure, or if maybe someone else is already working on
+it. You will save yourself a lot of time and effort by doing so.</p>
+</div>
+<div class="section" id="adding-a-new-intrinsic-function">
+<span id="intrinsic-function"></span><h2>Adding a new intrinsic function<a class="headerlink" href="#adding-a-new-intrinsic-function" title="Permalink to this headline">¶</a></h2>
+<p>Adding a new intrinsic function to LLVM is much easier than adding a new
+instruction.  Almost all extensions to LLVM should start as an intrinsic
+function and then be turned into an instruction if warranted.</p>
+<ol class="arabic">
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/docs/LangRef.html</span></tt>:</p>
+<p>Document the intrinsic.  Decide whether it is code generator specific and
+what the restrictions are.  Talk to other people about it so that you are
+sure it’s a good idea.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/include/llvm/IR/Intrinsics*.td</span></tt>:</p>
+<p>Add an entry for your intrinsic.  Describe its memory access characteristics
+for optimization (this controls whether it will be DCE’d, CSE’d, etc). Note
+that any intrinsic using the <tt class="docutils literal"><span class="pre">llvm_int_ty</span></tt> type for an argument will
+be deemed by <tt class="docutils literal"><span class="pre">tblgen</span></tt> as overloaded and the corresponding suffix will
+be required on the intrinsic’s name.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/Analysis/ConstantFolding.cpp</span></tt>:</p>
+<p>If it is possible to constant fold your intrinsic, add support to it in the
+<tt class="docutils literal"><span class="pre">canConstantFoldCallTo</span></tt> and <tt class="docutils literal"><span class="pre">ConstantFoldCall</span></tt> functions.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/test/Regression/*</span></tt>:</p>
+<p>Add test cases for your test cases to the test suite</p>
+</li>
+</ol>
+<p>Once the intrinsic has been added to the system, you must add code generator
+support for it.  Generally you must do the following steps:</p>
+<p>Add support to the .td file for the target(s) of your choice in
+<tt class="docutils literal"><span class="pre">lib/Target/*/*.td</span></tt>.</p>
+<blockquote>
+<div>This is usually a matter of adding a pattern to the .td file that matches the
+intrinsic, though it may obviously require adding the instructions you want to
+generate as well.  There are lots of examples in the PowerPC and X86 backend
+to follow.</div></blockquote>
+</div>
+<div class="section" id="adding-a-new-selectiondag-node">
+<h2>Adding a new SelectionDAG node<a class="headerlink" href="#adding-a-new-selectiondag-node" title="Permalink to this headline">¶</a></h2>
+<p>As with intrinsics, adding a new SelectionDAG node to LLVM is much easier than
+adding a new instruction.  New nodes are often added to help represent
+instructions common to many targets.  These nodes often map to an LLVM
+instruction (add, sub) or intrinsic (byteswap, population count).  In other
+cases, new nodes have been added to allow many targets to perform a common task
+(converting between floating point and integer representation) or capture more
+complicated behavior in a single node (rotate).</p>
+<ol class="arabic">
+<li><p class="first"><tt class="docutils literal"><span class="pre">include/llvm/CodeGen/ISDOpcodes.h</span></tt>:</p>
+<p>Add an enum value for the new SelectionDAG node.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">lib/CodeGen/SelectionDAG/SelectionDAG.cpp</span></tt>:</p>
+<dl class="docutils">
+<dt>Add code to print the node to <tt class="docutils literal"><span class="pre">getOperationName</span></tt>.  If your new node can be</dt>
+<dd><p class="first last">evaluated at compile time when given constant arguments (such as an add of a
+constant with another constant), find the <tt class="docutils literal"><span class="pre">getNode</span></tt> method that takes the
+appropriate number of arguments, and add a case for your node to the switch
+statement that performs constant folding for nodes that take the same number
+of arguments as your new node.</p>
+</dd>
+</dl>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">lib/CodeGen/SelectionDAG/LegalizeDAG.cpp</span></tt>:</p>
+<p>Add code to <a class="reference external" href="CodeGenerator.html#selectiondag_legalize">legalize, promote, and expand</a> the node as necessary.  At a
+minimum, you will need to add a case statement for your node in
+<tt class="docutils literal"><span class="pre">LegalizeOp</span></tt> which calls LegalizeOp on the node’s operands, and returns a
+new node if any of the operands changed as a result of being legalized.  It
+is likely that not all targets supported by the SelectionDAG framework will
+natively support the new node.  In this case, you must also add code in your
+node’s case statement in <tt class="docutils literal"><span class="pre">LegalizeOp</span></tt> to Expand your node into simpler,
+legal operations.  The case for <tt class="docutils literal"><span class="pre">ISD::UREM</span></tt> for expanding a remainder into
+a divide, multiply, and a subtract is a good example.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">lib/CodeGen/SelectionDAG/LegalizeDAG.cpp</span></tt>:</p>
+<dl class="docutils">
+<dt>If targets may support the new node being added only at certain sizes, you</dt>
+<dd><p class="first last">will also need to add code to your node’s case statement in <tt class="docutils literal"><span class="pre">LegalizeOp</span></tt>
+to Promote your node’s operands to a larger size, and perform the correct
+operation.  You will also need to add code to <tt class="docutils literal"><span class="pre">PromoteOp</span></tt> to do this as
+well.  For a good example, see <tt class="docutils literal"><span class="pre">ISD::BSWAP</span></tt>, which promotes its operand to
+a wider size, performs the byteswap, and then shifts the correct bytes right
+to emulate the narrower byteswap in the wider type.</p>
+</dd>
+</dl>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">lib/CodeGen/SelectionDAG/LegalizeDAG.cpp</span></tt>:</p>
+<p>Add a case for your node in <tt class="docutils literal"><span class="pre">ExpandOp</span></tt> to teach the legalizer how to
+perform the action represented by the new node on a value that has been split
+into high and low halves.  This case will be used to support your node with a
+64 bit operand on a 32 bit target.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">lib/CodeGen/SelectionDAG/DAGCombiner.cpp</span></tt>:</p>
+<p>If your node can be combined with itself, or other existing nodes in a
+peephole-like fashion, add a visit function for it, and call that function
+from. There are several good examples for simple combines you can do;
+<tt class="docutils literal"><span class="pre">visitFABS</span></tt> and <tt class="docutils literal"><span class="pre">visitSRL</span></tt> are good starting places.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">lib/Target/PowerPC/PPCISelLowering.cpp</span></tt>:</p>
+<p>Each target has an implementation of the <tt class="docutils literal"><span class="pre">TargetLowering</span></tt> class, usually in
+its own file (although some targets include it in the same file as the
+DAGToDAGISel).  The default behavior for a target is to assume that your new
+node is legal for all types that are legal for that target.  If this target
+does not natively support your node, then tell the target to either Promote
+it (if it is supported at a larger type) or Expand it.  This will cause the
+code you wrote in <tt class="docutils literal"><span class="pre">LegalizeOp</span></tt> above to decompose your new node into other
+legal nodes for this target.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">lib/Target/TargetSelectionDAG.td</span></tt>:</p>
+<p>Most current targets supported by LLVM generate code using the DAGToDAG
+method, where SelectionDAG nodes are pattern matched to target-specific
+nodes, which represent individual instructions.  In order for the targets to
+match an instruction to your new node, you must add a def for that node to
+the list in this file, with the appropriate type constraints. Look at
+<tt class="docutils literal"><span class="pre">add</span></tt>, <tt class="docutils literal"><span class="pre">bswap</span></tt>, and <tt class="docutils literal"><span class="pre">fadd</span></tt> for examples.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">lib/Target/PowerPC/PPCInstrInfo.td</span></tt>:</p>
+<p>Each target has a tablegen file that describes the target’s instruction set.
+For targets that use the DAGToDAG instruction selection framework, add a
+pattern for your new node that uses one or more target nodes.  Documentation
+for this is a bit sparse right now, but there are several decent examples.
+See the patterns for <tt class="docutils literal"><span class="pre">rotl</span></tt> in <tt class="docutils literal"><span class="pre">PPCInstrInfo.td</span></tt>.</p>
+</li>
+<li><p class="first">TODO: document complex patterns.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/test/Regression/CodeGen/*</span></tt>:</p>
+<p>Add test cases for your new node to the test suite.
+<tt class="docutils literal"><span class="pre">llvm/test/Regression/CodeGen/X86/bswap.ll</span></tt> is a good example.</p>
+</li>
+</ol>
+</div>
+<div class="section" id="adding-a-new-instruction">
+<h2>Adding a new instruction<a class="headerlink" href="#adding-a-new-instruction" title="Permalink to this headline">¶</a></h2>
+<div class="admonition warning">
+<p class="first admonition-title">Warning</p>
+<p class="last">Adding instructions changes the bitcode format, and it will take some effort
+to maintain compatibility with the previous version. Only add an instruction
+if it is absolutely necessary.</p>
+</div>
+<ol class="arabic">
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/include/llvm/Instruction.def</span></tt>:</p>
+<p>add a number for your instruction and an enum name</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/include/llvm/Instructions.h</span></tt>:</p>
+<p>add a definition for the class that will represent your instruction</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/include/llvm/Support/InstVisitor.h</span></tt>:</p>
+<p>add a prototype for a visitor to your new instruction type</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/AsmParser/Lexer.l</span></tt>:</p>
+<p>add a new token to parse your instruction from assembly text file</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/AsmParser/llvmAsmParser.y</span></tt>:</p>
+<p>add the grammar on how your instruction can be read and what it will
+construct as a result</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/Bitcode/Reader/Reader.cpp</span></tt>:</p>
+<p>add a case for your instruction and how it will be parsed from bitcode</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/VMCore/Instruction.cpp</span></tt>:</p>
+<p>add a case for how your instruction will be printed out to assembly</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/VMCore/Instructions.cpp</span></tt>:</p>
+<p>implement the class you defined in <tt class="docutils literal"><span class="pre">llvm/include/llvm/Instructions.h</span></tt></p>
+</li>
+<li><p class="first">Test your instruction</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/Target/*</span></tt>:</p>
+<p>add support for your instruction to code generators, or add a lowering pass.</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/test/Regression/*</span></tt>:</p>
+<p>add your test cases to the test suite.</p>
+</li>
+</ol>
+<p>Also, you need to implement (or modify) any analyses or passes that you want to
+understand this new instruction.</p>
+</div>
+<div class="section" id="adding-a-new-type">
+<h2>Adding a new type<a class="headerlink" href="#adding-a-new-type" title="Permalink to this headline">¶</a></h2>
+<div class="admonition warning">
+<p class="first admonition-title">Warning</p>
+<p class="last">Adding new types changes the bitcode format, and will break compatibility with
+currently-existing LLVM installations. Only add new types if it is absolutely
+necessary.</p>
+</div>
+<div class="section" id="adding-a-fundamental-type">
+<h3>Adding a fundamental type<a class="headerlink" href="#adding-a-fundamental-type" title="Permalink to this headline">¶</a></h3>
+<ol class="arabic">
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/include/llvm/Type.h</span></tt>:</p>
+<p>add enum for the new type; add static <tt class="docutils literal"><span class="pre">Type*</span></tt> for this type</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/VMCore/Type.cpp</span></tt>:</p>
+<p>add mapping from <tt class="docutils literal"><span class="pre">TypeID</span></tt> => <tt class="docutils literal"><span class="pre">Type*</span></tt>; initialize the static <tt class="docutils literal"><span class="pre">Type*</span></tt></p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/AsmReader/Lexer.l</span></tt>:</p>
+<p>add ability to parse in the type from text assembly</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/AsmReader/llvmAsmParser.y</span></tt>:</p>
+<p>add a token for that type</p>
+</li>
+</ol>
+</div>
+<div class="section" id="adding-a-derived-type">
+<h3>Adding a derived type<a class="headerlink" href="#adding-a-derived-type" title="Permalink to this headline">¶</a></h3>
+<ol class="arabic">
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/include/llvm/Type.h</span></tt>:</p>
+<p>add enum for the new type; add a forward declaration of the type also</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/include/llvm/DerivedTypes.h</span></tt>:</p>
+<p>add new class to represent new class in the hierarchy; add forward
+declaration to the TypeMap value type</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/VMCore/Type.cpp</span></tt>:</p>
+<p>add support for derived type to:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">std</span><span class="o">::</span><span class="n">string</span> <span class="n">getTypeDescription</span><span class="p">(</span><span class="k">const</span> <span class="n">Type</span> <span class="o">&</span><span class="n">Ty</span><span class="p">,</span>
+                               <span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o"><</span><span class="k">const</span> <span class="n">Type</span><span class="o">*></span> <span class="o">&</span><span class="n">TypeStack</span><span class="p">)</span>
+<span class="kt">bool</span> <span class="n">TypesEqual</span><span class="p">(</span><span class="k">const</span> <span class="n">Type</span> <span class="o">*</span><span class="n">Ty</span><span class="p">,</span> <span class="k">const</span> <span class="n">Type</span> <span class="o">*</span><span class="n">Ty2</span><span class="p">,</span>
+                <span class="n">std</span><span class="o">::</span><span class="n">map</span><span class="o"><</span><span class="k">const</span> <span class="n">Type</span><span class="o">*</span><span class="p">,</span> <span class="k">const</span> <span class="n">Type</span><span class="o">*></span> <span class="o">&</span><span class="n">EqTypes</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>add necessary member functions for type, and factory methods</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/AsmReader/Lexer.l</span></tt>:</p>
+<p>add ability to parse in the type from text assembly</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/Bitcode/Writer/Writer.cpp</span></tt>:</p>
+<p>modify <tt class="docutils literal"><span class="pre">void</span> <span class="pre">BitcodeWriter::outputType(const</span> <span class="pre">Type</span> <span class="pre">*T)</span></tt> to serialize your
+type</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/Bitcode/Reader/Reader.cpp</span></tt>:</p>
+<p>modify <tt class="docutils literal"><span class="pre">const</span> <span class="pre">Type</span> <span class="pre">*BitcodeReader::ParseType()</span></tt> to read your data type</p>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">llvm/lib/VMCore/AsmWriter.cpp</span></tt>:</p>
+<p>modify</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="kt">void</span> <span class="n">calcTypeName</span><span class="p">(</span><span class="k">const</span> <span class="n">Type</span> <span class="o">*</span><span class="n">Ty</span><span class="p">,</span>
+                  <span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o"><</span><span class="k">const</span> <span class="n">Type</span><span class="o">*></span> <span class="o">&</span><span class="n">TypeStack</span><span class="p">,</span>
+                  <span class="n">std</span><span class="o">::</span><span class="n">map</span><span class="o"><</span><span class="k">const</span> <span class="n">Type</span><span class="o">*</span><span class="p">,</span><span class="n">std</span><span class="o">::</span><span class="n">string</span><span class="o">></span> <span class="o">&</span><span class="n">TypeNames</span><span class="p">,</span>
+                  <span class="n">std</span><span class="o">::</span><span class="n">string</span> <span class="o">&</span><span class="n">Result</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>to output the new derived type</p>
+</li>
+</ol>
+</div>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="HowToSetUpLLVMStyleRTTI.html" title="How to set up LLVM-style RTTI for your class hierarchy"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="CompilerWriterInfo.html" title="Architecture & Platform Information for Compiler Writers"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/Extensions.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/Extensions.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/Extensions.html (added)
+++ www-releases/trunk/3.5.1/docs/Extensions.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,296 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>LLVM Extensions — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="LLVM Alias Analysis Infrastructure" href="AliasAnalysis.html" />
+    <link rel="prev" title="LLVM Programmer’s Manual" href="ProgrammersManual.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">
+      <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="AliasAnalysis.html" title="LLVM Alias Analysis Infrastructure"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="ProgrammersManual.html" title="LLVM Programmer’s Manual"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="llvm-extensions">
+<h1>LLVM Extensions<a class="headerlink" href="#llvm-extensions" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#introduction" id="id1">Introduction</a></li>
+<li><a class="reference internal" href="#general-assembly-syntax" id="id2">General Assembly Syntax</a><ul>
+<li><a class="reference internal" href="#c99-style-hexadecimal-floating-point-constants" id="id3">C99-style Hexadecimal Floating-point Constants</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#machine-specific-assembly-syntax" id="id4">Machine-specific Assembly Syntax</a><ul>
+<li><a class="reference internal" href="#x86-coff-dependent" id="id5">X86/COFF-Dependent</a><ul>
+<li><a class="reference internal" href="#relocations" id="id6">Relocations</a></li>
+<li><a class="reference internal" href="#linkonce-directive" id="id7"><tt class="docutils literal"><span class="pre">.linkonce</span></tt> Directive</a></li>
+<li><a class="reference internal" href="#section-directive" id="id8"><tt class="docutils literal"><span class="pre">.section</span></tt> Directive</a></li>
+</ul>
+</li>
+</ul>
+</li>
+<li><a class="reference internal" href="#target-specific-behaviour" id="id9">Target Specific Behaviour</a><ul>
+<li><a class="reference internal" href="#windows-on-arm" id="id10">Windows on ARM</a><ul>
+<li><a class="reference internal" href="#stack-probe-emission" id="id11">Stack Probe Emission</a></li>
+<li><a class="reference internal" href="#variable-length-arrays" id="id12">Variable Length Arrays</a></li>
+</ul>
+</li>
+</ul>
+</li>
+</ul>
+</div>
+<div class="toctree-wrapper compound">
+</div>
+<div class="section" id="introduction">
+<h2><a class="toc-backref" href="#id1">Introduction</a><a class="headerlink" href="#introduction" title="Permalink to this headline">¶</a></h2>
+<p>This document describes extensions to tools and formats LLVM seeks compatibility
+with.</p>
+</div>
+<div class="section" id="general-assembly-syntax">
+<h2><a class="toc-backref" href="#id2">General Assembly Syntax</a><a class="headerlink" href="#general-assembly-syntax" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="c99-style-hexadecimal-floating-point-constants">
+<h3><a class="toc-backref" href="#id3">C99-style Hexadecimal Floating-point Constants</a><a class="headerlink" href="#c99-style-hexadecimal-floating-point-constants" title="Permalink to this headline">¶</a></h3>
+<p>LLVM’s assemblers allow floating-point constants to be written in C99’s
+hexadecimal format instead of decimal if desired.</p>
+<div class="highlight-gas"><div class="highlight"><pre><span class="na">.section</span> <span class="no">.data</span>
+<span class="na">.float</span> <span class="mi">0x1c2</span><span class="no">.2ap3</span>
+</pre></div>
+</div>
+</div>
+</div>
+<div class="section" id="machine-specific-assembly-syntax">
+<h2><a class="toc-backref" href="#id4">Machine-specific Assembly Syntax</a><a class="headerlink" href="#machine-specific-assembly-syntax" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="x86-coff-dependent">
+<h3><a class="toc-backref" href="#id5">X86/COFF-Dependent</a><a class="headerlink" href="#x86-coff-dependent" title="Permalink to this headline">¶</a></h3>
+<div class="section" id="relocations">
+<h4><a class="toc-backref" href="#id6">Relocations</a><a class="headerlink" href="#relocations" title="Permalink to this headline">¶</a></h4>
+<p>The following additional relocation types are supported:</p>
+<p><strong>@IMGREL</strong> (AT&T syntax only) generates an image-relative relocation that
+corresponds to the COFF relocation types <tt class="docutils literal"><span class="pre">IMAGE_REL_I386_DIR32NB</span></tt> (32-bit) or
+<tt class="docutils literal"><span class="pre">IMAGE_REL_AMD64_ADDR32NB</span></tt> (64-bit).</p>
+<div class="highlight-gas"><pre>.text
+fun:
+  mov foo at IMGREL(%ebx, %ecx, 4), %eax
+
+.section .pdata
+  .long fun at IMGREL
+  .long (fun at imgrel + 0x3F)
+  .long $unwind$fun at imgrel</pre>
+</div>
+<p><strong>.secrel32</strong> generates a relocation that corresponds to the COFF relocation
+types <tt class="docutils literal"><span class="pre">IMAGE_REL_I386_SECREL</span></tt> (32-bit) or <tt class="docutils literal"><span class="pre">IMAGE_REL_AMD64_SECREL</span></tt> (64-bit).</p>
+<p><strong>.secidx</strong> relocation generates an index of the section that contains
+the target.  It corresponds to the COFF relocation types
+<tt class="docutils literal"><span class="pre">IMAGE_REL_I386_SECTION</span></tt> (32-bit) or <tt class="docutils literal"><span class="pre">IMAGE_REL_AMD64_SECTION</span></tt> (64-bit).</p>
+<div class="highlight-gas"><div class="highlight"><pre><span class="na">.section</span> <span class="no">.debug$S</span><span class="p">,</span><span class="s">"rn"</span>
+  <span class="na">.long</span> <span class="mi">4</span>
+  <span class="na">.long</span> <span class="mi">242</span>
+  <span class="na">.long</span> <span class="mi">40</span>
+  <span class="na">.secrel32</span> <span class="no">_function_name</span>
+  <span class="na">.secidx</span>   <span class="no">_function_name</span>
+  <span class="na">...</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="linkonce-directive">
+<h4><a class="toc-backref" href="#id7"><tt class="docutils literal"><span class="pre">.linkonce</span></tt> Directive</a><a class="headerlink" href="#linkonce-directive" title="Permalink to this headline">¶</a></h4>
+<p>Syntax:</p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">.linkonce</span> <span class="pre">[</span> <span class="pre">comdat</span> <span class="pre">type</span> <span class="pre">]</span></tt></div></blockquote>
+<p>Supported COMDAT types:</p>
+<dl class="docutils">
+<dt><tt class="docutils literal"><span class="pre">discard</span></tt></dt>
+<dd>Discards duplicate sections with the same COMDAT symbol. This is the default
+if no type is specified.</dd>
+<dt><tt class="docutils literal"><span class="pre">one_only</span></tt></dt>
+<dd>If the symbol is defined multiple times, the linker issues an error.</dd>
+<dt><tt class="docutils literal"><span class="pre">same_size</span></tt></dt>
+<dd>Duplicates are discarded, but the linker issues an error if any have
+different sizes.</dd>
+<dt><tt class="docutils literal"><span class="pre">same_contents</span></tt></dt>
+<dd>Duplicates are discarded, but the linker issues an error if any duplicates
+do not have exactly the same content.</dd>
+<dt><tt class="docutils literal"><span class="pre">largest</span></tt></dt>
+<dd>Links the largest section from among the duplicates.</dd>
+<dt><tt class="docutils literal"><span class="pre">newest</span></tt></dt>
+<dd>Links the newest section from among the duplicates.</dd>
+</dl>
+<div class="highlight-gas"><div class="highlight"><pre><span class="na">.section</span> <span class="no">.text$foo</span>
+<span class="na">.linkonce</span>
+  <span class="na">...</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="section-directive">
+<h4><a class="toc-backref" href="#id8"><tt class="docutils literal"><span class="pre">.section</span></tt> Directive</a><a class="headerlink" href="#section-directive" title="Permalink to this headline">¶</a></h4>
+<p>MC supports passing the information in <tt class="docutils literal"><span class="pre">.linkonce</span></tt> at the end of
+<tt class="docutils literal"><span class="pre">.section</span></tt>. For example,  these two codes are equivalent</p>
+<div class="highlight-gas"><div class="highlight"><pre><span class="na">.section</span> <span class="no">secName</span><span class="p">,</span> <span class="s">"dr"</span><span class="p">,</span> <span class="no">discard</span><span class="p">,</span> <span class="s">"Symbol1"</span>
+<span class="na">.globl</span> <span class="no">Symbol1</span>
+<span class="nl">Symbol1:</span>
+<span class="na">.long</span> <span class="mi">1</span>
+</pre></div>
+</div>
+<div class="highlight-gas"><div class="highlight"><pre><span class="na">.section</span> <span class="no">secName</span><span class="p">,</span> <span class="s">"dr"</span>
+<span class="na">.linkonce</span> <span class="no">discard</span>
+<span class="na">.globl</span> <span class="no">Symbol1</span>
+<span class="nl">Symbol1:</span>
+<span class="na">.long</span> <span class="mi">1</span>
+</pre></div>
+</div>
+<p>Note that in the combined form the COMDAT symbol is explicit. This
+extension exists to support multiple sections with the same name in
+different COMDATs:</p>
+<div class="highlight-gas"><div class="highlight"><pre><span class="na">.section</span> <span class="no">secName</span><span class="p">,</span> <span class="s">"dr"</span><span class="p">,</span> <span class="no">discard</span><span class="p">,</span> <span class="s">"Symbol1"</span>
+<span class="na">.globl</span> <span class="no">Symbol1</span>
+<span class="nl">Symbol1:</span>
+<span class="na">.long</span> <span class="mi">1</span>
+
+<span class="na">.section</span> <span class="no">secName</span><span class="p">,</span> <span class="s">"dr"</span><span class="p">,</span> <span class="no">discard</span><span class="p">,</span> <span class="s">"Symbol2"</span>
+<span class="na">.globl</span> <span class="no">Symbol2</span>
+<span class="nl">Symbol2:</span>
+<span class="na">.long</span> <span class="mi">1</span>
+</pre></div>
+</div>
+<p>In addition to the types allowed with <tt class="docutils literal"><span class="pre">.linkonce</span></tt>, <tt class="docutils literal"><span class="pre">.section</span></tt> also accepts
+<tt class="docutils literal"><span class="pre">associative</span></tt>. The meaning is that the section is linked  if a certain other
+COMDAT section is linked. This other section is indicated by the comdat symbol
+in this directive. It can be any symbol defined in the associated section, but
+is usually the associated section’s comdat.</p>
+<blockquote>
+<div><p>The following restrictions apply to the associated section:</p>
+<ol class="arabic simple">
+<li>It must be a COMDAT section.</li>
+<li>It cannot be another associative COMDAT section.</li>
+</ol>
+</div></blockquote>
+<p>In the following example the symobl <tt class="docutils literal"><span class="pre">sym</span></tt> is the comdat symbol of <tt class="docutils literal"><span class="pre">.foo</span></tt>
+and <tt class="docutils literal"><span class="pre">.bar</span></tt> is associated to <tt class="docutils literal"><span class="pre">.foo</span></tt>.</p>
+<div class="highlight-gas"><div class="highlight"><pre><span class="na">.section</span>        <span class="no">.foo</span><span class="p">,</span><span class="s">"bw"</span><span class="p">,</span><span class="no">discard</span><span class="p">,</span> <span class="s">"sym"</span>
+<span class="na">.section</span>        <span class="no">.bar</span><span class="p">,</span><span class="s">"rd"</span><span class="p">,</span><span class="no">associative</span><span class="p">,</span> <span class="s">"sym"</span>
+</pre></div>
+</div>
+</div>
+</div>
+</div>
+<div class="section" id="target-specific-behaviour">
+<h2><a class="toc-backref" href="#id9">Target Specific Behaviour</a><a class="headerlink" href="#target-specific-behaviour" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="windows-on-arm">
+<h3><a class="toc-backref" href="#id10">Windows on ARM</a><a class="headerlink" href="#windows-on-arm" title="Permalink to this headline">¶</a></h3>
+<div class="section" id="stack-probe-emission">
+<h4><a class="toc-backref" href="#id11">Stack Probe Emission</a><a class="headerlink" href="#stack-probe-emission" title="Permalink to this headline">¶</a></h4>
+<p>The reference implementation (Microsoft Visual Studio 2012) emits stack probes
+in the following fashion:</p>
+<div class="highlight-gas"><div class="highlight"><pre><span class="nf">movw</span> <span class="no">r4</span><span class="p">,</span> <span class="c">#constant</span>
+<span class="nf">bl</span> <span class="no">__chkstk</span>
+<span class="nf">sub.w</span> <span class="no">sp</span><span class="p">,</span> <span class="no">sp</span><span class="p">,</span> <span class="no">r4</span>
+</pre></div>
+</div>
+<p>However, this has the limitation of 32 MiB (±16MiB).  In order to accommodate
+larger binaries, LLVM supports the use of <tt class="docutils literal"><span class="pre">-mcode-model=large</span></tt> to allow a 4GiB
+range via a slight deviation.  It will generate an indirect jump as follows:</p>
+<div class="highlight-gas"><div class="highlight"><pre><span class="nf">movw</span> <span class="no">r4</span><span class="p">,</span> <span class="c">#constant</span>
+<span class="nf">movw</span> <span class="no">r12</span><span class="p">,</span> <span class="p">:</span><span class="no">lower16</span><span class="p">:</span><span class="no">__chkstk</span>
+<span class="nf">movt</span> <span class="no">r12</span><span class="p">,</span> <span class="p">:</span><span class="no">upper16</span><span class="p">:</span><span class="no">__chkstk</span>
+<span class="nf">blx</span> <span class="no">r12</span>
+<span class="nf">sub.w</span> <span class="no">sp</span><span class="p">,</span> <span class="no">sp</span><span class="p">,</span> <span class="no">r4</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="variable-length-arrays">
+<h4><a class="toc-backref" href="#id12">Variable Length Arrays</a><a class="headerlink" href="#variable-length-arrays" title="Permalink to this headline">¶</a></h4>
+<p>The reference implementation (Microsoft Visual Studio 2012) does not permit the
+emission of Variable Length Arrays (VLAs).</p>
+<p>The Windows ARM Itanium ABI extends the base ABI by adding support for emitting
+a dynamic stack allocation.  When emitting a variable stack allocation, a call
+to <tt class="docutils literal"><span class="pre">__chkstk</span></tt> is emitted unconditionally to ensure that guard pages are setup
+properly.  The emission of this stack probe emission is handled similar to the
+standard stack probe emission.</p>
+<p>The MSVC environment does not emit code for VLAs currently.</p>
+</div>
+</div>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="AliasAnalysis.html" title="LLVM Alias Analysis Infrastructure"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="ProgrammersManual.html" title="LLVM Programmer’s Manual"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/FAQ.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/FAQ.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/FAQ.html (added)
+++ www-releases/trunk/3.5.1/docs/FAQ.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,557 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>Frequently Asked Questions (FAQ) — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="The LLVM Lexicon" href="Lexicon.html" />
+    <link rel="prev" title="Getting Started with the LLVM System using Microsoft Visual Studio" href="GettingStartedVS.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">
+      <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="Lexicon.html" title="The LLVM Lexicon"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="GettingStartedVS.html" title="Getting Started with the LLVM System using Microsoft Visual Studio"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="frequently-asked-questions-faq">
+<h1>Frequently Asked Questions (FAQ)<a class="headerlink" href="#frequently-asked-questions-faq" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#license" id="id1">License</a><ul>
+<li><a class="reference internal" href="#does-the-university-of-illinois-open-source-license-really-qualify-as-an-open-source-license" id="id2">Does the University of Illinois Open Source License really qualify as an “open source” license?</a></li>
+<li><a class="reference internal" href="#can-i-modify-llvm-source-code-and-redistribute-the-modified-source" id="id3">Can I modify LLVM source code and redistribute the modified source?</a></li>
+<li><a class="reference internal" href="#can-i-modify-the-llvm-source-code-and-redistribute-binaries-or-other-tools-based-on-it-without-redistributing-the-source" id="id4">Can I modify the LLVM source code and redistribute binaries or other tools based on it, without redistributing the source?</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#source-code" id="id5">Source Code</a><ul>
+<li><a class="reference internal" href="#in-what-language-is-llvm-written" id="id6">In what language is LLVM written?</a></li>
+<li><a class="reference internal" href="#how-portable-is-the-llvm-source-code" id="id7">How portable is the LLVM source code?</a></li>
+<li><a class="reference internal" href="#what-api-do-i-use-to-store-a-value-to-one-of-the-virtual-registers-in-llvm-ir-s-ssa-representation" id="id8">What API do I use to store a value to one of the virtual registers in LLVM IR’s SSA representation?</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#build-problems" id="id9">Build Problems</a><ul>
+<li><a class="reference internal" href="#when-i-run-configure-it-finds-the-wrong-c-compiler" id="id10">When I run configure, it finds the wrong C compiler.</a></li>
+<li><a class="reference internal" href="#the-configure-script-finds-the-right-c-compiler-but-it-uses-the-llvm-tools-from-a-previous-build-what-do-i-do" id="id11">The <tt class="docutils literal"><span class="pre">configure</span></tt> script finds the right C compiler, but it uses the LLVM tools from a previous build.  What do I do?</a></li>
+<li><a class="reference internal" href="#when-creating-a-dynamic-library-i-get-a-strange-glibc-error" id="id12">When creating a dynamic library, I get a strange GLIBC error.</a></li>
+<li><a class="reference internal" href="#i-ve-updated-my-source-tree-from-subversion-and-now-my-build-is-trying-to-use-a-file-directory-that-doesn-t-exist" id="id13">I’ve updated my source tree from Subversion, and now my build is trying to use a file/directory that doesn’t exist.</a></li>
+<li><a class="reference internal" href="#i-ve-modified-a-makefile-in-my-source-tree-but-my-build-tree-keeps-using-the-old-version-what-do-i-do" id="id14">I’ve modified a Makefile in my source tree, but my build tree keeps using the old version.  What do I do?</a></li>
+<li><a class="reference internal" href="#i-ve-upgraded-to-a-new-version-of-llvm-and-i-get-strange-build-errors" id="id15">I’ve upgraded to a new version of LLVM, and I get strange build errors.</a></li>
+<li><a class="reference internal" href="#i-ve-built-llvm-and-am-testing-it-but-the-tests-freeze" id="id16">I’ve built LLVM and am testing it, but the tests freeze.</a></li>
+<li><a class="reference internal" href="#why-do-test-results-differ-when-i-perform-different-types-of-builds" id="id17">Why do test results differ when I perform different types of builds?</a></li>
+<li><a class="reference internal" href="#compiling-llvm-with-gcc-3-3-2-fails-what-should-i-do" id="id18">Compiling LLVM with GCC 3.3.2 fails, what should I do?</a></li>
+<li><a class="reference internal" href="#after-subversion-update-rebuilding-gives-the-error-no-rule-to-make-target" id="id19">After Subversion update, rebuilding gives the error “No rule to make target”.</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#source-languages" id="id20">Source Languages</a><ul>
+<li><a class="reference internal" href="#what-source-languages-are-supported" id="id21">What source languages are supported?</a></li>
+<li><a class="reference internal" href="#i-d-like-to-write-a-self-hosting-llvm-compiler-how-should-i-interface-with-the-llvm-middle-end-optimizers-and-back-end-code-generators" id="id22">I’d like to write a self-hosting LLVM compiler. How should I interface with the LLVM middle-end optimizers and back-end code generators?</a></li>
+<li><a class="reference internal" href="#what-support-is-there-for-a-higher-level-source-language-constructs-for-building-a-compiler" id="id23">What support is there for a higher level source language constructs for building a compiler?</a></li>
+<li><a class="reference internal" href="#i-don-t-understand-the-getelementptr-instruction-help" id="id24">I don’t understand the <tt class="docutils literal"><span class="pre">GetElementPtr</span></tt> instruction. Help!</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#using-the-c-and-c-front-ends" id="id25">Using the C and C++ Front Ends</a><ul>
+<li><a class="reference internal" href="#can-i-compile-c-or-c-code-to-platform-independent-llvm-bitcode" id="id26">Can I compile C or C++ code to platform-independent LLVM bitcode?</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#questions-about-code-generated-by-the-demo-page" id="id27">Questions about code generated by the demo page</a><ul>
+<li><a class="reference internal" href="#what-is-this-llvm-global-ctors-and-global-i-a-stuff-that-happens-when-i-include-iostream" id="id28">What is this <tt class="docutils literal"><span class="pre">llvm.global_ctors</span></tt> and <tt class="docutils literal"><span class="pre">_GLOBAL__I_a...</span></tt> stuff that happens when I <tt class="docutils literal"><span class="pre">#include</span> <span class="pre"><iostream></span></tt>?</a></li>
+<li><a class="reference internal" href="#where-did-all-of-my-code-go" id="id29">Where did all of my code go??</a></li>
+<li><a class="reference internal" href="#what-is-this-undef-thing-that-shows-up-in-my-code" id="id30">What is this “<tt class="docutils literal"><span class="pre">undef</span></tt>” thing that shows up in my code?</a></li>
+<li><a class="reference internal" href="#why-does-instcombine-simplifycfg-turn-a-call-to-a-function-with-a-mismatched-calling-convention-into-unreachable-why-not-make-the-verifier-reject-it" id="id31">Why does instcombine + simplifycfg turn a call to a function with a mismatched calling convention into “unreachable”? Why not make the verifier reject it?</a></li>
+</ul>
+</li>
+</ul>
+</div>
+<div class="section" id="license">
+<h2><a class="toc-backref" href="#id1">License</a><a class="headerlink" href="#license" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="does-the-university-of-illinois-open-source-license-really-qualify-as-an-open-source-license">
+<h3><a class="toc-backref" href="#id2">Does the University of Illinois Open Source License really qualify as an “open source” license?</a><a class="headerlink" href="#does-the-university-of-illinois-open-source-license-really-qualify-as-an-open-source-license" title="Permalink to this headline">¶</a></h3>
+<p>Yes, the license is <a class="reference external" href="http://www.opensource.org/licenses/UoI-NCSA.php">certified</a> by the Open Source
+Initiative (OSI).</p>
+</div>
+<div class="section" id="can-i-modify-llvm-source-code-and-redistribute-the-modified-source">
+<h3><a class="toc-backref" href="#id3">Can I modify LLVM source code and redistribute the modified source?</a><a class="headerlink" href="#can-i-modify-llvm-source-code-and-redistribute-the-modified-source" title="Permalink to this headline">¶</a></h3>
+<p>Yes.  The modified source distribution must retain the copyright notice and
+follow the three bulletted conditions listed in the <a class="reference external" href="http://llvm.org/svn/llvm-project/llvm/trunk/LICENSE.TXT">LLVM license</a>.</p>
+</div>
+<div class="section" id="can-i-modify-the-llvm-source-code-and-redistribute-binaries-or-other-tools-based-on-it-without-redistributing-the-source">
+<h3><a class="toc-backref" href="#id4">Can I modify the LLVM source code and redistribute binaries or other tools based on it, without redistributing the source?</a><a class="headerlink" href="#can-i-modify-the-llvm-source-code-and-redistribute-binaries-or-other-tools-based-on-it-without-redistributing-the-source" title="Permalink to this headline">¶</a></h3>
+<p>Yes. This is why we distribute LLVM under a less restrictive license than GPL,
+as explained in the first question above.</p>
+</div>
+</div>
+<div class="section" id="source-code">
+<h2><a class="toc-backref" href="#id5">Source Code</a><a class="headerlink" href="#source-code" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="in-what-language-is-llvm-written">
+<h3><a class="toc-backref" href="#id6">In what language is LLVM written?</a><a class="headerlink" href="#in-what-language-is-llvm-written" title="Permalink to this headline">¶</a></h3>
+<p>All of the LLVM tools and libraries are written in C++ with extensive use of
+the STL.</p>
+</div>
+<div class="section" id="how-portable-is-the-llvm-source-code">
+<h3><a class="toc-backref" href="#id7">How portable is the LLVM source code?</a><a class="headerlink" href="#how-portable-is-the-llvm-source-code" title="Permalink to this headline">¶</a></h3>
+<p>The LLVM source code should be portable to most modern Unix-like operating
+systems.  Most of the code is written in standard C++ with operating system
+services abstracted to a support library.  The tools required to build and
+test LLVM have been ported to a plethora of platforms.</p>
+<p>Some porting problems may exist in the following areas:</p>
+<ul class="simple">
+<li>The autoconf/makefile build system relies heavily on UNIX shell tools,
+like the Bourne Shell and sed.  Porting to systems without these tools
+(MacOS 9, Plan 9) will require more effort.</li>
+</ul>
+</div>
+<div class="section" id="what-api-do-i-use-to-store-a-value-to-one-of-the-virtual-registers-in-llvm-ir-s-ssa-representation">
+<h3><a class="toc-backref" href="#id8">What API do I use to store a value to one of the virtual registers in LLVM IR’s SSA representation?</a><a class="headerlink" href="#what-api-do-i-use-to-store-a-value-to-one-of-the-virtual-registers-in-llvm-ir-s-ssa-representation" title="Permalink to this headline">¶</a></h3>
+<p>In short: you can’t. It’s actually kind of a silly question once you grok
+what’s going on. Basically, in code like:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="nv">%result</span> <span class="p">=</span> <span class="k">add</span> <span class="k">i32</span> <span class="nv">%foo</span><span class="p">,</span> <span class="nv">%bar</span>
+</pre></div>
+</div>
+<p>, <tt class="docutils literal"><span class="pre">%result</span></tt> is just a name given to the <tt class="docutils literal"><span class="pre">Value</span></tt> of the <tt class="docutils literal"><span class="pre">add</span></tt>
+instruction. In other words, <tt class="docutils literal"><span class="pre">%result</span></tt> <em>is</em> the add instruction. The
+“assignment” doesn’t explicitly “store” anything to any “virtual register”;
+the “<tt class="docutils literal"><span class="pre">=</span></tt>” is more like the mathematical sense of equality.</p>
+<p>Longer explanation: In order to generate a textual representation of the
+IR, some kind of name has to be given to each instruction so that other
+instructions can textually reference it. However, the isomorphic in-memory
+representation that you manipulate from C++ has no such restriction since
+instructions can simply keep pointers to any other <tt class="docutils literal"><span class="pre">Value</span></tt>‘s that they
+reference. In fact, the names of dummy numbered temporaries like <tt class="docutils literal"><span class="pre">%1</span></tt> are
+not explicitly represented in the in-memory representation at all (see
+<tt class="docutils literal"><span class="pre">Value::getName()</span></tt>).</p>
+</div>
+</div>
+<div class="section" id="build-problems">
+<h2><a class="toc-backref" href="#id9">Build Problems</a><a class="headerlink" href="#build-problems" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="when-i-run-configure-it-finds-the-wrong-c-compiler">
+<h3><a class="toc-backref" href="#id10">When I run configure, it finds the wrong C compiler.</a><a class="headerlink" href="#when-i-run-configure-it-finds-the-wrong-c-compiler" title="Permalink to this headline">¶</a></h3>
+<p>The <tt class="docutils literal"><span class="pre">configure</span></tt> script attempts to locate first <tt class="docutils literal"><span class="pre">gcc</span></tt> and then <tt class="docutils literal"><span class="pre">cc</span></tt>,
+unless it finds compiler paths set in <tt class="docutils literal"><span class="pre">CC</span></tt> and <tt class="docutils literal"><span class="pre">CXX</span></tt> for the C and C++
+compiler, respectively.</p>
+<p>If <tt class="docutils literal"><span class="pre">configure</span></tt> finds the wrong compiler, either adjust your <tt class="docutils literal"><span class="pre">PATH</span></tt>
+environment variable or set <tt class="docutils literal"><span class="pre">CC</span></tt> and <tt class="docutils literal"><span class="pre">CXX</span></tt> explicitly.</p>
+</div>
+<div class="section" id="the-configure-script-finds-the-right-c-compiler-but-it-uses-the-llvm-tools-from-a-previous-build-what-do-i-do">
+<h3><a class="toc-backref" href="#id11">The <tt class="docutils literal"><span class="pre">configure</span></tt> script finds the right C compiler, but it uses the LLVM tools from a previous build.  What do I do?</a><a class="headerlink" href="#the-configure-script-finds-the-right-c-compiler-but-it-uses-the-llvm-tools-from-a-previous-build-what-do-i-do" title="Permalink to this headline">¶</a></h3>
+<p>The <tt class="docutils literal"><span class="pre">configure</span></tt> script uses the <tt class="docutils literal"><span class="pre">PATH</span></tt> to find executables, so if it’s
+grabbing the wrong linker/assembler/etc, there are two ways to fix it:</p>
+<ol class="arabic simple">
+<li>Adjust your <tt class="docutils literal"><span class="pre">PATH</span></tt> environment variable so that the correct program
+appears first in the <tt class="docutils literal"><span class="pre">PATH</span></tt>.  This may work, but may not be convenient
+when you want them <em>first</em> in your path for other work.</li>
+<li>Run <tt class="docutils literal"><span class="pre">configure</span></tt> with an alternative <tt class="docutils literal"><span class="pre">PATH</span></tt> that is correct. In a
+Bourne compatible shell, the syntax would be:</li>
+</ol>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nv">PATH</span><span class="o">=[</span>the path without the bad program<span class="o">]</span> ./configure ...
+</pre></div>
+</div>
+<p>This is still somewhat inconvenient, but it allows <tt class="docutils literal"><span class="pre">configure</span></tt> to do its
+work without having to adjust your <tt class="docutils literal"><span class="pre">PATH</span></tt> permanently.</p>
+</div>
+<div class="section" id="when-creating-a-dynamic-library-i-get-a-strange-glibc-error">
+<h3><a class="toc-backref" href="#id12">When creating a dynamic library, I get a strange GLIBC error.</a><a class="headerlink" href="#when-creating-a-dynamic-library-i-get-a-strange-glibc-error" title="Permalink to this headline">¶</a></h3>
+<p>Under some operating systems (i.e. Linux), libtool does not work correctly if
+GCC was compiled with the <tt class="docutils literal"><span class="pre">--disable-shared</span> <span class="pre">option</span></tt>.  To work around this,
+install your own version of GCC that has shared libraries enabled by default.</p>
+</div>
+<div class="section" id="i-ve-updated-my-source-tree-from-subversion-and-now-my-build-is-trying-to-use-a-file-directory-that-doesn-t-exist">
+<h3><a class="toc-backref" href="#id13">I’ve updated my source tree from Subversion, and now my build is trying to use a file/directory that doesn’t exist.</a><a class="headerlink" href="#i-ve-updated-my-source-tree-from-subversion-and-now-my-build-is-trying-to-use-a-file-directory-that-doesn-t-exist" title="Permalink to this headline">¶</a></h3>
+<p>You need to re-run configure in your object directory.  When new Makefiles
+are added to the source tree, they have to be copied over to the object tree
+in order to be used by the build.</p>
+</div>
+<div class="section" id="i-ve-modified-a-makefile-in-my-source-tree-but-my-build-tree-keeps-using-the-old-version-what-do-i-do">
+<h3><a class="toc-backref" href="#id14">I’ve modified a Makefile in my source tree, but my build tree keeps using the old version.  What do I do?</a><a class="headerlink" href="#i-ve-modified-a-makefile-in-my-source-tree-but-my-build-tree-keeps-using-the-old-version-what-do-i-do" title="Permalink to this headline">¶</a></h3>
+<p>If the Makefile already exists in your object tree, you can just run the
+following command in the top level directory of your object tree:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> ./config.status <relative path to Makefile>;
+</pre></div>
+</div>
+<p>If the Makefile is new, you will have to modify the configure script to copy
+it over.</p>
+</div>
+<div class="section" id="i-ve-upgraded-to-a-new-version-of-llvm-and-i-get-strange-build-errors">
+<h3><a class="toc-backref" href="#id15">I’ve upgraded to a new version of LLVM, and I get strange build errors.</a><a class="headerlink" href="#i-ve-upgraded-to-a-new-version-of-llvm-and-i-get-strange-build-errors" title="Permalink to this headline">¶</a></h3>
+<p>Sometimes, changes to the LLVM source code alters how the build system works.
+Changes in <tt class="docutils literal"><span class="pre">libtool</span></tt>, <tt class="docutils literal"><span class="pre">autoconf</span></tt>, or header file dependencies are
+especially prone to this sort of problem.</p>
+<p>The best thing to try is to remove the old files and re-build.  In most cases,
+this takes care of the problem.  To do this, just type <tt class="docutils literal"><span class="pre">make</span> <span class="pre">clean</span></tt> and then
+<tt class="docutils literal"><span class="pre">make</span></tt> in the directory that fails to build.</p>
+</div>
+<div class="section" id="i-ve-built-llvm-and-am-testing-it-but-the-tests-freeze">
+<h3><a class="toc-backref" href="#id16">I’ve built LLVM and am testing it, but the tests freeze.</a><a class="headerlink" href="#i-ve-built-llvm-and-am-testing-it-but-the-tests-freeze" title="Permalink to this headline">¶</a></h3>
+<p>This is most likely occurring because you built a profile or release
+(optimized) build of LLVM and have not specified the same information on the
+<tt class="docutils literal"><span class="pre">gmake</span></tt> command line.</p>
+<p>For example, if you built LLVM with the command:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> gmake <span class="nv">ENABLE_PROFILING</span><span class="o">=</span>1
+</pre></div>
+</div>
+<p>...then you must run the tests with the following commands:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nb">cd </span>llvm/test
+<span class="gp">%</span> gmake <span class="nv">ENABLE_PROFILING</span><span class="o">=</span>1
+</pre></div>
+</div>
+</div>
+<div class="section" id="why-do-test-results-differ-when-i-perform-different-types-of-builds">
+<h3><a class="toc-backref" href="#id17">Why do test results differ when I perform different types of builds?</a><a class="headerlink" href="#why-do-test-results-differ-when-i-perform-different-types-of-builds" title="Permalink to this headline">¶</a></h3>
+<p>The LLVM test suite is dependent upon several features of the LLVM tools and
+libraries.</p>
+<p>First, the debugging assertions in code are not enabled in optimized or
+profiling builds.  Hence, tests that used to fail may pass.</p>
+<p>Second, some tests may rely upon debugging options or behavior that is only
+available in the debug build.  These tests will fail in an optimized or
+profile build.</p>
+</div>
+<div class="section" id="compiling-llvm-with-gcc-3-3-2-fails-what-should-i-do">
+<h3><a class="toc-backref" href="#id18">Compiling LLVM with GCC 3.3.2 fails, what should I do?</a><a class="headerlink" href="#compiling-llvm-with-gcc-3-3-2-fails-what-should-i-do" title="Permalink to this headline">¶</a></h3>
+<p>This is <a class="reference external" href="http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13392">a bug in GCC</a>,
+and affects projects other than LLVM.  Try upgrading or downgrading your GCC.</p>
+</div>
+<div class="section" id="after-subversion-update-rebuilding-gives-the-error-no-rule-to-make-target">
+<h3><a class="toc-backref" href="#id19">After Subversion update, rebuilding gives the error “No rule to make target”.</a><a class="headerlink" href="#after-subversion-update-rebuilding-gives-the-error-no-rule-to-make-target" title="Permalink to this headline">¶</a></h3>
+<p>If the error is of the form:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="go">gmake[2]: *** No rule to make target `/path/to/somefile',</span>
+<span class="go">              needed by `/path/to/another/file.d'.</span>
+<span class="go">Stop.</span>
+</pre></div>
+</div>
+<p>This may occur anytime files are moved within the Subversion repository or
+removed entirely.  In this case, the best solution is to erase all <tt class="docutils literal"><span class="pre">.d</span></tt>
+files, which list dependencies for source files, and rebuild:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nb">cd</span> <span class="nv">$LLVM_OBJ_DIR</span>
+<span class="gp">%</span> rm -f <span class="sb">`</span>find . -name <span class="se">\*\.</span>d<span class="sb">`</span>
+<span class="gp">%</span> gmake
+</pre></div>
+</div>
+<p>In other cases, it may be necessary to run <tt class="docutils literal"><span class="pre">make</span> <span class="pre">clean</span></tt> before rebuilding.</p>
+</div>
+</div>
+<div class="section" id="source-languages">
+<h2><a class="toc-backref" href="#id20">Source Languages</a><a class="headerlink" href="#source-languages" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="what-source-languages-are-supported">
+<h3><a class="toc-backref" href="#id21">What source languages are supported?</a><a class="headerlink" href="#what-source-languages-are-supported" title="Permalink to this headline">¶</a></h3>
+<p>LLVM currently has full support for C and C++ source languages. These are
+available through both <a class="reference external" href="http://clang.llvm.org/">Clang</a> and <a class="reference external" href="http://dragonegg.llvm.org/">DragonEgg</a>.</p>
+<p>The PyPy developers are working on integrating LLVM into the PyPy backend so
+that PyPy language can translate to LLVM.</p>
+</div>
+<div class="section" id="i-d-like-to-write-a-self-hosting-llvm-compiler-how-should-i-interface-with-the-llvm-middle-end-optimizers-and-back-end-code-generators">
+<h3><a class="toc-backref" href="#id22">I’d like to write a self-hosting LLVM compiler. How should I interface with the LLVM middle-end optimizers and back-end code generators?</a><a class="headerlink" href="#i-d-like-to-write-a-self-hosting-llvm-compiler-how-should-i-interface-with-the-llvm-middle-end-optimizers-and-back-end-code-generators" title="Permalink to this headline">¶</a></h3>
+<p>Your compiler front-end will communicate with LLVM by creating a module in the
+LLVM intermediate representation (IR) format. Assuming you want to write your
+language’s compiler in the language itself (rather than C++), there are 3
+major ways to tackle generating LLVM IR from a front-end:</p>
+<ol class="arabic simple">
+<li><strong>Call into the LLVM libraries code using your language’s FFI (foreign
+function interface).</strong></li>
+</ol>
+<blockquote>
+<div><ul class="simple">
+<li><em>for:</em> best tracks changes to the LLVM IR, .ll syntax, and .bc format</li>
+<li><em>for:</em> enables running LLVM optimization passes without a emit/parse
+overhead</li>
+<li><em>for:</em> adapts well to a JIT context</li>
+<li><em>against:</em> lots of ugly glue code to write</li>
+</ul>
+</div></blockquote>
+<ol class="arabic simple" start="2">
+<li><strong>Emit LLVM assembly from your compiler’s native language.</strong></li>
+</ol>
+<blockquote>
+<div><ul class="simple">
+<li><em>for:</em> very straightforward to get started</li>
+<li><em>against:</em> the .ll parser is slower than the bitcode reader when
+interfacing to the middle end</li>
+<li><em>against:</em> it may be harder to track changes to the IR</li>
+</ul>
+</div></blockquote>
+<ol class="arabic simple" start="3">
+<li><strong>Emit LLVM bitcode from your compiler’s native language.</strong></li>
+</ol>
+<blockquote>
+<div><ul class="simple">
+<li><em>for:</em> can use the more-efficient bitcode reader when interfacing to the
+middle end</li>
+<li><em>against:</em> you’ll have to re-engineer the LLVM IR object model and bitcode
+writer in your language</li>
+<li><em>against:</em> it may be harder to track changes to the IR</li>
+</ul>
+</div></blockquote>
+<p>If you go with the first option, the C bindings in include/llvm-c should help
+a lot, since most languages have strong support for interfacing with C. The
+most common hurdle with calling C from managed code is interfacing with the
+garbage collector. The C interface was designed to require very little memory
+management, and so is straightforward in this regard.</p>
+</div>
+<div class="section" id="what-support-is-there-for-a-higher-level-source-language-constructs-for-building-a-compiler">
+<h3><a class="toc-backref" href="#id23">What support is there for a higher level source language constructs for building a compiler?</a><a class="headerlink" href="#what-support-is-there-for-a-higher-level-source-language-constructs-for-building-a-compiler" title="Permalink to this headline">¶</a></h3>
+<p>Currently, there isn’t much. LLVM supports an intermediate representation
+which is useful for code representation but will not support the high level
+(abstract syntax tree) representation needed by most compilers. There are no
+facilities for lexical nor semantic analysis.</p>
+</div>
+<div class="section" id="i-don-t-understand-the-getelementptr-instruction-help">
+<h3><a class="toc-backref" href="#id24">I don’t understand the <tt class="docutils literal"><span class="pre">GetElementPtr</span></tt> instruction. Help!</a><a class="headerlink" href="#i-don-t-understand-the-getelementptr-instruction-help" title="Permalink to this headline">¶</a></h3>
+<p>See <a class="reference external" href="GetElementPtr.html">The Often Misunderstood GEP Instruction</a>.</p>
+</div>
+</div>
+<div class="section" id="using-the-c-and-c-front-ends">
+<h2><a class="toc-backref" href="#id25">Using the C and C++ Front Ends</a><a class="headerlink" href="#using-the-c-and-c-front-ends" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="can-i-compile-c-or-c-code-to-platform-independent-llvm-bitcode">
+<h3><a class="toc-backref" href="#id26">Can I compile C or C++ code to platform-independent LLVM bitcode?</a><a class="headerlink" href="#can-i-compile-c-or-c-code-to-platform-independent-llvm-bitcode" title="Permalink to this headline">¶</a></h3>
+<p>No. C and C++ are inherently platform-dependent languages. The most obvious
+example of this is the preprocessor. A very common way that C code is made
+portable is by using the preprocessor to include platform-specific code. In
+practice, information about other platforms is lost after preprocessing, so
+the result is inherently dependent on the platform that the preprocessing was
+targeting.</p>
+<p>Another example is <tt class="docutils literal"><span class="pre">sizeof</span></tt>. It’s common for <tt class="docutils literal"><span class="pre">sizeof(long)</span></tt> to vary
+between platforms. In most C front-ends, <tt class="docutils literal"><span class="pre">sizeof</span></tt> is expanded to a
+constant immediately, thus hard-wiring a platform-specific detail.</p>
+<p>Also, since many platforms define their ABIs in terms of C, and since LLVM is
+lower-level than C, front-ends currently must emit platform-specific IR in
+order to have the result conform to the platform ABI.</p>
+</div>
+</div>
+<div class="section" id="questions-about-code-generated-by-the-demo-page">
+<h2><a class="toc-backref" href="#id27">Questions about code generated by the demo page</a><a class="headerlink" href="#questions-about-code-generated-by-the-demo-page" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="what-is-this-llvm-global-ctors-and-global-i-a-stuff-that-happens-when-i-include-iostream">
+<h3><a class="toc-backref" href="#id28">What is this <tt class="docutils literal"><span class="pre">llvm.global_ctors</span></tt> and <tt class="docutils literal"><span class="pre">_GLOBAL__I_a...</span></tt> stuff that happens when I <tt class="docutils literal"><span class="pre">#include</span> <span class="pre"><iostream></span></tt>?</a><a class="headerlink" href="#what-is-this-llvm-global-ctors-and-global-i-a-stuff-that-happens-when-i-include-iostream" title="Permalink to this headline">¶</a></h3>
+<p>If you <tt class="docutils literal"><span class="pre">#include</span></tt> the <tt class="docutils literal"><span class="pre"><iostream></span></tt> header into a C++ translation unit,
+the file will probably use the <tt class="docutils literal"><span class="pre">std::cin</span></tt>/<tt class="docutils literal"><span class="pre">std::cout</span></tt>/... global objects.
+However, C++ does not guarantee an order of initialization between static
+objects in different translation units, so if a static ctor/dtor in your .cpp
+file used <tt class="docutils literal"><span class="pre">std::cout</span></tt>, for example, the object would not necessarily be
+automatically initialized before your use.</p>
+<p>To make <tt class="docutils literal"><span class="pre">std::cout</span></tt> and friends work correctly in these scenarios, the STL
+that we use declares a static object that gets created in every translation
+unit that includes <tt class="docutils literal"><span class="pre"><iostream></span></tt>.  This object has a static constructor
+and destructor that initializes and destroys the global iostream objects
+before they could possibly be used in the file.  The code that you see in the
+<tt class="docutils literal"><span class="pre">.ll</span></tt> file corresponds to the constructor and destructor registration code.</p>
+<p>If you would like to make it easier to <em>understand</em> the LLVM code generated
+by the compiler in the demo page, consider using <tt class="docutils literal"><span class="pre">printf()</span></tt> instead of
+<tt class="docutils literal"><span class="pre">iostream</span></tt>s to print values.</p>
+</div>
+<div class="section" id="where-did-all-of-my-code-go">
+<h3><a class="toc-backref" href="#id29">Where did all of my code go??</a><a class="headerlink" href="#where-did-all-of-my-code-go" title="Permalink to this headline">¶</a></h3>
+<p>If you are using the LLVM demo page, you may often wonder what happened to
+all of the code that you typed in.  Remember that the demo script is running
+the code through the LLVM optimizers, so if your code doesn’t actually do
+anything useful, it might all be deleted.</p>
+<p>To prevent this, make sure that the code is actually needed.  For example, if
+you are computing some expression, return the value from the function instead
+of leaving it in a local variable.  If you really want to constrain the
+optimizer, you can read from and assign to <tt class="docutils literal"><span class="pre">volatile</span></tt> global variables.</p>
+</div>
+<div class="section" id="what-is-this-undef-thing-that-shows-up-in-my-code">
+<h3><a class="toc-backref" href="#id30">What is this “<tt class="docutils literal"><span class="pre">undef</span></tt>” thing that shows up in my code?</a><a class="headerlink" href="#what-is-this-undef-thing-that-shows-up-in-my-code" title="Permalink to this headline">¶</a></h3>
+<p><tt class="docutils literal"><span class="pre">undef</span></tt> is the LLVM way of representing a value that is not defined.  You
+can get these if you do not initialize a variable before you use it.  For
+example, the C function:</p>
+<div class="highlight-c"><div class="highlight"><pre><span class="kt">int</span> <span class="nf">X</span><span class="p">()</span> <span class="p">{</span> <span class="kt">int</span> <span class="n">i</span><span class="p">;</span> <span class="k">return</span> <span class="n">i</span><span class="p">;</span> <span class="p">}</span>
+</pre></div>
+</div>
+<p>Is compiled to “<tt class="docutils literal"><span class="pre">ret</span> <span class="pre">i32</span> <span class="pre">undef</span></tt>” because “<tt class="docutils literal"><span class="pre">i</span></tt>” never has a value specified
+for it.</p>
+</div>
+<div class="section" id="why-does-instcombine-simplifycfg-turn-a-call-to-a-function-with-a-mismatched-calling-convention-into-unreachable-why-not-make-the-verifier-reject-it">
+<h3><a class="toc-backref" href="#id31">Why does instcombine + simplifycfg turn a call to a function with a mismatched calling convention into “unreachable”? Why not make the verifier reject it?</a><a class="headerlink" href="#why-does-instcombine-simplifycfg-turn-a-call-to-a-function-with-a-mismatched-calling-convention-into-unreachable-why-not-make-the-verifier-reject-it" title="Permalink to this headline">¶</a></h3>
+<p>This is a common problem run into by authors of front-ends that are using
+custom calling conventions: you need to make sure to set the right calling
+convention on both the function and on each call to the function.  For
+example, this code:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">define</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span> <span class="p">{</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+<span class="k">define</span> <span class="kt">void</span> <span class="vg">@bar</span><span class="p">()</span> <span class="p">{</span>
+    <span class="k">call</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>Is optimized to:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">define</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span> <span class="p">{</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+<span class="k">define</span> <span class="kt">void</span> <span class="vg">@bar</span><span class="p">()</span> <span class="p">{</span>
+    <span class="k">unreachable</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>... with “<tt class="docutils literal"><span class="pre">opt</span> <span class="pre">-instcombine</span> <span class="pre">-simplifycfg</span></tt>”.  This often bites people because
+“all their code disappears”.  Setting the calling convention on the caller and
+callee is required for indirect calls to work, so people often ask why not
+make the verifier reject this sort of thing.</p>
+<p>The answer is that this code has undefined behavior, but it is not illegal.
+If we made it illegal, then every transformation that could potentially create
+this would have to ensure that it doesn’t, and there is valid code that can
+create this sort of construct (in dead code).  The sorts of things that can
+cause this to happen are fairly contrived, but we still need to accept them.
+Here’s an example:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">define</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span> <span class="p">{</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+<span class="k">define</span> <span class="k">internal</span> <span class="kt">void</span> <span class="vg">@bar</span><span class="p">(</span><span class="kt">void</span><span class="p">()*</span> <span class="nv">%FP</span><span class="p">,</span> <span class="k">i1</span> <span class="nv">%cond</span><span class="p">)</span> <span class="p">{</span>
+    <span class="k">br</span> <span class="k">i1</span> <span class="nv">%cond</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%T</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%F</span>
+<span class="nl">T:</span>
+    <span class="k">call</span> <span class="kt">void</span> <span class="nv">%FP</span><span class="p">()</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="nl">F:</span>
+    <span class="k">call</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="nv">%FP</span><span class="p">()</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+<span class="k">define</span> <span class="kt">void</span> <span class="vg">@test</span><span class="p">()</span> <span class="p">{</span>
+    <span class="nv">%X</span> <span class="p">=</span> <span class="k">or</span> <span class="k">i1</span> <span class="k">false</span><span class="p">,</span> <span class="k">false</span>
+    <span class="k">call</span> <span class="kt">void</span> <span class="vg">@bar</span><span class="p">(</span><span class="kt">void</span><span class="p">()*</span> <span class="vg">@foo</span><span class="p">,</span> <span class="k">i1</span> <span class="nv">%X</span><span class="p">)</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>In this example, “test” always passes <tt class="docutils literal"><span class="pre">@foo</span></tt>/<tt class="docutils literal"><span class="pre">false</span></tt> into <tt class="docutils literal"><span class="pre">bar</span></tt>, which
+ensures that it is dynamically called with the right calling conv (thus, the
+code is perfectly well defined).  If you run this through the inliner, you
+get this (the explicit “or” is there so that the inliner doesn’t dead code
+eliminate a bunch of stuff):</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">define</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span> <span class="p">{</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+<span class="k">define</span> <span class="kt">void</span> <span class="vg">@test</span><span class="p">()</span> <span class="p">{</span>
+    <span class="nv">%X</span> <span class="p">=</span> <span class="k">or</span> <span class="k">i1</span> <span class="k">false</span><span class="p">,</span> <span class="k">false</span>
+    <span class="k">br</span> <span class="k">i1</span> <span class="nv">%X</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%T.i</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%F.i</span>
+<span class="nl">T.i:</span>
+    <span class="k">call</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span>
+    <span class="k">br</span> <span class="kt">label</span> <span class="nv">%bar.exit</span>
+<span class="nl">F.i:</span>
+    <span class="k">call</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span>
+    <span class="k">br</span> <span class="kt">label</span> <span class="nv">%bar.exit</span>
+<span class="nl">bar.exit:</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>Here you can see that the inlining pass made an undefined call to <tt class="docutils literal"><span class="pre">@foo</span></tt>
+with the wrong calling convention.  We really don’t want to make the inliner
+have to know about this sort of thing, so it needs to be valid code.  In this
+case, dead code elimination can trivially remove the undefined code.  However,
+if <tt class="docutils literal"><span class="pre">%X</span></tt> was an input argument to <tt class="docutils literal"><span class="pre">@test</span></tt>, the inliner would produce this:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">define</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span> <span class="p">{</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+
+<span class="k">define</span> <span class="kt">void</span> <span class="vg">@test</span><span class="p">(</span><span class="k">i1</span> <span class="nv">%X</span><span class="p">)</span> <span class="p">{</span>
+    <span class="k">br</span> <span class="k">i1</span> <span class="nv">%X</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%T.i</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%F.i</span>
+<span class="nl">T.i:</span>
+    <span class="k">call</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span>
+    <span class="k">br</span> <span class="kt">label</span> <span class="nv">%bar.exit</span>
+<span class="nl">F.i:</span>
+    <span class="k">call</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span>
+    <span class="k">br</span> <span class="kt">label</span> <span class="nv">%bar.exit</span>
+<span class="nl">bar.exit:</span>
+    <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>The interesting thing about this is that <tt class="docutils literal"><span class="pre">%X</span></tt> <em>must</em> be false for the
+code to be well-defined, but no amount of dead code elimination will be able
+to delete the broken call as unreachable.  However, since
+<tt class="docutils literal"><span class="pre">instcombine</span></tt>/<tt class="docutils literal"><span class="pre">simplifycfg</span></tt> turns the undefined call into unreachable, we
+end up with a branch on a condition that goes to unreachable: a branch to
+unreachable can never happen, so “<tt class="docutils literal"><span class="pre">-inline</span> <span class="pre">-instcombine</span> <span class="pre">-simplifycfg</span></tt>” is
+able to produce:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">define</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span> <span class="p">{</span>
+   <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+<span class="k">define</span> <span class="kt">void</span> <span class="vg">@test</span><span class="p">(</span><span class="k">i1</span> <span class="nv">%X</span><span class="p">)</span> <span class="p">{</span>
+<span class="nl">F.i:</span>
+   <span class="k">call</span> <span class="k">fastcc</span> <span class="kt">void</span> <span class="vg">@foo</span><span class="p">()</span>
+   <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+</div>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="Lexicon.html" title="The LLVM Lexicon"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="GettingStartedVS.html" title="Getting Started with the LLVM System using Microsoft Visual Studio"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/GarbageCollection.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/GarbageCollection.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/GarbageCollection.html (added)
+++ www-releases/trunk/3.5.1/docs/GarbageCollection.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,1151 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>Accurate Garbage Collection with LLVM — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="Writing an LLVM Pass" href="WritingAnLLVMPass.html" />
+    <link rel="prev" title="How To Use Instruction Mappings" href="HowToUseInstrMappings.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">
+      <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="WritingAnLLVMPass.html" title="Writing an LLVM Pass"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="HowToUseInstrMappings.html" title="How To Use Instruction Mappings"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="accurate-garbage-collection-with-llvm">
+<h1>Accurate Garbage Collection with LLVM<a class="headerlink" href="#accurate-garbage-collection-with-llvm" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#introduction" id="id1">Introduction</a><ul>
+<li><a class="reference internal" href="#goals-and-non-goals" id="id2">Goals and non-goals</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#getting-started" id="id3">Getting started</a><ul>
+<li><a class="reference internal" href="#in-your-compiler" id="id4">In your compiler</a></li>
+<li><a class="reference internal" href="#in-your-runtime" id="id5">In your runtime</a></li>
+<li><a class="reference internal" href="#about-the-shadow-stack" id="id6">About the shadow stack</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#ir-features" id="id7">IR features</a><ul>
+<li><a class="reference internal" href="#specifying-gc-code-generation-gc" id="id8">Specifying GC code generation: <tt class="docutils literal"><span class="pre">gc</span> <span class="pre">"..."</span></tt></a></li>
+<li><a class="reference internal" href="#gcroot" id="id9">Identifying GC roots on the stack: <tt class="docutils literal"><span class="pre">llvm.gcroot</span></tt></a></li>
+<li><a class="reference internal" href="#reading-and-writing-references-in-the-heap" id="id10">Reading and writing references in the heap</a><ul>
+<li><a class="reference internal" href="#write-barrier-llvm-gcwrite" id="id11">Write barrier: <tt class="docutils literal"><span class="pre">llvm.gcwrite</span></tt></a></li>
+<li><a class="reference internal" href="#read-barrier-llvm-gcread" id="id12">Read barrier: <tt class="docutils literal"><span class="pre">llvm.gcread</span></tt></a></li>
+</ul>
+</li>
+</ul>
+</li>
+<li><a class="reference internal" href="#implementing-a-collector-plugin" id="id13">Implementing a collector plugin</a><ul>
+<li><a class="reference internal" href="#overview-of-available-features" id="id14">Overview of available features</a></li>
+<li><a class="reference internal" href="#computing-stack-maps" id="id15">Computing stack maps</a></li>
+<li><a class="reference internal" href="#initializing-roots-to-null-initroots" id="id16">Initializing roots to null: <tt class="docutils literal"><span class="pre">InitRoots</span></tt></a></li>
+<li><a class="reference internal" href="#custom-lowering-of-intrinsics-customroots-customreadbarriers-and-customwritebarriers" id="id17">Custom lowering of intrinsics: <tt class="docutils literal"><span class="pre">CustomRoots</span></tt>, <tt class="docutils literal"><span class="pre">CustomReadBarriers</span></tt>, and <tt class="docutils literal"><span class="pre">CustomWriteBarriers</span></tt></a></li>
+<li><a class="reference internal" href="#generating-safe-points-neededsafepoints" id="id18">Generating safe points: <tt class="docutils literal"><span class="pre">NeededSafePoints</span></tt></a></li>
+<li><a class="reference internal" href="#emitting-assembly-code-gcmetadataprinter" id="id19">Emitting assembly code: <tt class="docutils literal"><span class="pre">GCMetadataPrinter</span></tt></a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#references" id="id20">References</a></li>
+</ul>
+</div>
+<div class="section" id="introduction">
+<h2><a class="toc-backref" href="#id1">Introduction</a><a class="headerlink" href="#introduction" title="Permalink to this headline">¶</a></h2>
+<p>Garbage collection is a widely used technique that frees the programmer from
+having to know the lifetimes of heap objects, making software easier to produce
+and maintain.  Many programming languages rely on garbage collection for
+automatic memory management.  There are two primary forms of garbage collection:
+conservative and accurate.</p>
+<p>Conservative garbage collection often does not require any special support from
+either the language or the compiler: it can handle non-type-safe programming
+languages (such as C/C++) and does not require any special information from the
+compiler.  The <a class="reference external" href="http://www.hpl.hp.com/personal/Hans_Boehm/gc/">Boehm collector</a> is an example of a
+state-of-the-art conservative collector.</p>
+<p>Accurate garbage collection requires the ability to identify all pointers in the
+program at run-time (which requires that the source-language be type-safe in
+most cases).  Identifying pointers at run-time requires compiler support to
+locate all places that hold live pointer variables at run-time, including the
+<a class="reference internal" href="#gcroot"><em>processor stack and registers</em></a>.</p>
+<p>Conservative garbage collection is attractive because it does not require any
+special compiler support, but it does have problems.  In particular, because the
+conservative garbage collector cannot <em>know</em> that a particular word in the
+machine is a pointer, it cannot move live objects in the heap (preventing the
+use of compacting and generational GC algorithms) and it can occasionally suffer
+from memory leaks due to integer values that happen to point to objects in the
+program.  In addition, some aggressive compiler transformations can break
+conservative garbage collectors (though these seem rare in practice).</p>
+<p>Accurate garbage collectors do not suffer from any of these problems, but they
+can suffer from degraded scalar optimization of the program.  In particular,
+because the runtime must be able to identify and update all pointers active in
+the program, some optimizations are less effective.  In practice, however, the
+locality and performance benefits of using aggressive garbage collection
+techniques dominates any low-level losses.</p>
+<p>This document describes the mechanisms and interfaces provided by LLVM to
+support accurate garbage collection.</p>
+<div class="section" id="goals-and-non-goals">
+<h3><a class="toc-backref" href="#id2">Goals and non-goals</a><a class="headerlink" href="#goals-and-non-goals" title="Permalink to this headline">¶</a></h3>
+<p>LLVM’s intermediate representation provides <a class="reference internal" href="#gc-intrinsics"><em>garbage collection intrinsics</em></a> that offer support for a broad class of collector models.  For
+instance, the intrinsics permit:</p>
+<ul class="simple">
+<li>semi-space collectors</li>
+<li>mark-sweep collectors</li>
+<li>generational collectors</li>
+<li>reference counting</li>
+<li>incremental collectors</li>
+<li>concurrent collectors</li>
+<li>cooperative collectors</li>
+</ul>
+<p>We hope that the primitive support built into the LLVM IR is sufficient to
+support a broad class of garbage collected languages including Scheme, ML, Java,
+C#, Perl, Python, Lua, Ruby, other scripting languages, and more.</p>
+<p>However, LLVM does not itself provide a garbage collector — this should be
+part of your language’s runtime library.  LLVM provides a framework for compile
+time <a class="reference internal" href="#plugin"><em>code generation plugins</em></a>.  The role of these plugins is to
+generate code and data structures which conforms to the <em>binary interface</em>
+specified by the <em>runtime library</em>.  This is similar to the relationship between
+LLVM and DWARF debugging info, for example.  The difference primarily lies in
+the lack of an established standard in the domain of garbage collection — thus
+the plugins.</p>
+<p>The aspects of the binary interface with which LLVM’s GC support is
+concerned are:</p>
+<ul class="simple">
+<li>Creation of GC-safe points within code where collection is allowed to execute
+safely.</li>
+<li>Computation of the stack map.  For each safe point in the code, object
+references within the stack frame must be identified so that the collector may
+traverse and perhaps update them.</li>
+<li>Write barriers when storing object references to the heap.  These are commonly
+used to optimize incremental scans in generational collectors.</li>
+<li>Emission of read barriers when loading object references.  These are useful
+for interoperating with concurrent collectors.</li>
+</ul>
+<p>There are additional areas that LLVM does not directly address:</p>
+<ul class="simple">
+<li>Registration of global roots with the runtime.</li>
+<li>Registration of stack map entries with the runtime.</li>
+<li>The functions used by the program to allocate memory, trigger a collection,
+etc.</li>
+<li>Computation or compilation of type maps, or registration of them with the
+runtime.  These are used to crawl the heap for object references.</li>
+</ul>
+<p>In general, LLVM’s support for GC does not include features which can be
+adequately addressed with other features of the IR and does not specify a
+particular binary interface.  On the plus side, this means that you should be
+able to integrate LLVM with an existing runtime.  On the other hand, it leaves a
+lot of work for the developer of a novel language.  However, it’s easy to get
+started quickly and scale up to a more sophisticated implementation as your
+compiler matures.</p>
+</div>
+</div>
+<div class="section" id="getting-started">
+<h2><a class="toc-backref" href="#id3">Getting started</a><a class="headerlink" href="#getting-started" title="Permalink to this headline">¶</a></h2>
+<p>Using a GC with LLVM implies many things, for example:</p>
+<ul class="simple">
+<li>Write a runtime library or find an existing one which implements a GC heap.<ol class="arabic">
+<li>Implement a memory allocator.</li>
+<li>Design a binary interface for the stack map, used to identify references
+within a stack frame on the machine stack.*</li>
+<li>Implement a stack crawler to discover functions on the call stack.*</li>
+<li>Implement a registry for global roots.</li>
+<li>Design a binary interface for type maps, used to identify references
+within heap objects.</li>
+<li>Implement a collection routine bringing together all of the above.</li>
+</ol>
+</li>
+<li>Emit compatible code from your compiler.<ul>
+<li>Initialization in the main function.</li>
+<li>Use the <tt class="docutils literal"><span class="pre">gc</span> <span class="pre">"..."</span></tt> attribute to enable GC code generation (or
+<tt class="docutils literal"><span class="pre">F.setGC("...")</span></tt>).</li>
+<li>Use <tt class="docutils literal"><span class="pre">@llvm.gcroot</span></tt> to mark stack roots.</li>
+<li>Use <tt class="docutils literal"><span class="pre">@llvm.gcread</span></tt> and/or <tt class="docutils literal"><span class="pre">@llvm.gcwrite</span></tt> to manipulate GC references,
+if necessary.</li>
+<li>Allocate memory using the GC allocation routine provided by the runtime
+library.</li>
+<li>Generate type maps according to your runtime’s binary interface.</li>
+</ul>
+</li>
+<li>Write a compiler plugin to interface LLVM with the runtime library.*<ul>
+<li>Lower <tt class="docutils literal"><span class="pre">@llvm.gcread</span></tt> and <tt class="docutils literal"><span class="pre">@llvm.gcwrite</span></tt> to appropriate code
+sequences.*</li>
+<li>Compile LLVM’s stack map to the binary form expected by the runtime.</li>
+</ul>
+</li>
+<li>Load the plugin into the compiler.  Use <tt class="docutils literal"><span class="pre">llc</span> <span class="pre">-load</span></tt> or link the plugin
+statically with your language’s compiler.*</li>
+<li>Link program executables with the runtime.</li>
+</ul>
+<p>To help with several of these tasks (those indicated with a *), LLVM includes a
+highly portable, built-in ShadowStack code generator.  It is compiled into
+<tt class="docutils literal"><span class="pre">llc</span></tt> and works even with the interpreter and C backends.</p>
+<div class="section" id="in-your-compiler">
+<h3><a class="toc-backref" href="#id4">In your compiler</a><a class="headerlink" href="#in-your-compiler" title="Permalink to this headline">¶</a></h3>
+<p>To turn the shadow stack on for your functions, first call:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">F</span><span class="p">.</span><span class="n">setGC</span><span class="p">(</span><span class="s">"shadow-stack"</span><span class="p">);</span>
+</pre></div>
+</div>
+<p>for each function your compiler emits. Since the shadow stack is built into
+LLVM, you do not need to load a plugin.</p>
+<p>Your compiler must also use <tt class="docutils literal"><span class="pre">@llvm.gcroot</span></tt> as documented.  Don’t forget to
+create a root for each intermediate value that is generated when evaluating an
+expression.  In <tt class="docutils literal"><span class="pre">h(f(),</span> <span class="pre">g())</span></tt>, the result of <tt class="docutils literal"><span class="pre">f()</span></tt> could easily be collected
+if evaluating <tt class="docutils literal"><span class="pre">g()</span></tt> triggers a collection.</p>
+<p>There’s no need to use <tt class="docutils literal"><span class="pre">@llvm.gcread</span></tt> and <tt class="docutils literal"><span class="pre">@llvm.gcwrite</span></tt> over plain
+<tt class="docutils literal"><span class="pre">load</span></tt> and <tt class="docutils literal"><span class="pre">store</span></tt> for now.  You will need them when switching to a more
+advanced GC.</p>
+</div>
+<div class="section" id="in-your-runtime">
+<h3><a class="toc-backref" href="#id5">In your runtime</a><a class="headerlink" href="#in-your-runtime" title="Permalink to this headline">¶</a></h3>
+<p>The shadow stack doesn’t imply a memory allocation algorithm.  A semispace
+collector or building atop <tt class="docutils literal"><span class="pre">malloc</span></tt> are great places to start, and can be
+implemented with very little code.</p>
+<p>When it comes time to collect, however, your runtime needs to traverse the stack
+roots, and for this it needs to integrate with the shadow stack.  Luckily, doing
+so is very simple. (This code is heavily commented to help you understand the
+data structure, but there are only 20 lines of meaningful code.)</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="c1">/// @brief The map for a single function's stack frame.  One of these is</span>
+<span class="c1">///        compiled as constant data into the executable for each function.</span>
+<span class="c1">///</span>
+<span class="c1">/// Storage of metadata values is elided if the %metadata parameter to</span>
+<span class="c1">/// @llvm.gcroot is null.</span>
+<span class="k">struct</span> <span class="n">FrameMap</span> <span class="p">{</span>
+  <span class="kt">int32_t</span> <span class="n">NumRoots</span><span class="p">;</span>    <span class="c1">//< Number of roots in stack frame.</span>
+  <span class="kt">int32_t</span> <span class="n">NumMeta</span><span class="p">;</span>     <span class="c1">//< Number of metadata entries.  May be < NumRoots.</span>
+  <span class="k">const</span> <span class="kt">void</span> <span class="o">*</span><span class="n">Meta</span><span class="p">[</span><span class="mi">0</span><span class="p">];</span> <span class="c1">//< Metadata for each root.</span>
+<span class="p">};</span>
+
+<span class="c1">/// @brief A link in the dynamic shadow stack.  One of these is embedded in</span>
+<span class="c1">///        the stack frame of each function on the call stack.</span>
+<span class="k">struct</span> <span class="n">StackEntry</span> <span class="p">{</span>
+  <span class="n">StackEntry</span> <span class="o">*</span><span class="n">Next</span><span class="p">;</span>    <span class="c1">//< Link to next stack entry (the caller's).</span>
+  <span class="k">const</span> <span class="n">FrameMap</span> <span class="o">*</span><span class="n">Map</span><span class="p">;</span> <span class="c1">//< Pointer to constant FrameMap.</span>
+  <span class="kt">void</span> <span class="o">*</span><span class="n">Roots</span><span class="p">[</span><span class="mi">0</span><span class="p">];</span>      <span class="c1">//< Stack roots (in-place array).</span>
+<span class="p">};</span>
+
+<span class="c1">/// @brief The head of the singly-linked list of StackEntries.  Functions push</span>
+<span class="c1">///        and pop onto this in their prologue and epilogue.</span>
+<span class="c1">///</span>
+<span class="c1">/// Since there is only a global list, this technique is not threadsafe.</span>
+<span class="n">StackEntry</span> <span class="o">*</span><span class="n">llvm_gc_root_chain</span><span class="p">;</span>
+
+<span class="c1">/// @brief Calls Visitor(root, meta) for each GC root on the stack.</span>
+<span class="c1">///        root and meta are exactly the values passed to</span>
+<span class="c1">///        @llvm.gcroot.</span>
+<span class="c1">///</span>
+<span class="c1">/// Visitor could be a function to recursively mark live objects.  Or it</span>
+<span class="c1">/// might copy them to another heap or generation.</span>
+<span class="c1">///</span>
+<span class="c1">/// @param Visitor A function to invoke for every GC root on the stack.</span>
+<span class="kt">void</span> <span class="nf">visitGCRoots</span><span class="p">(</span><span class="kt">void</span> <span class="p">(</span><span class="o">*</span><span class="n">Visitor</span><span class="p">)(</span><span class="kt">void</span> <span class="o">**</span><span class="n">Root</span><span class="p">,</span> <span class="k">const</span> <span class="kt">void</span> <span class="o">*</span><span class="n">Meta</span><span class="p">))</span> <span class="p">{</span>
+  <span class="k">for</span> <span class="p">(</span><span class="n">StackEntry</span> <span class="o">*</span><span class="n">R</span> <span class="o">=</span> <span class="n">llvm_gc_root_chain</span><span class="p">;</span> <span class="n">R</span><span class="p">;</span> <span class="n">R</span> <span class="o">=</span> <span class="n">R</span><span class="o">-></span><span class="n">Next</span><span class="p">)</span> <span class="p">{</span>
+    <span class="kt">unsigned</span> <span class="n">i</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span>
+
+    <span class="c1">// For roots [0, NumMeta), the metadata pointer is in the FrameMap.</span>
+    <span class="k">for</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="n">e</span> <span class="o">=</span> <span class="n">R</span><span class="o">-></span><span class="n">Map</span><span class="o">-></span><span class="n">NumMeta</span><span class="p">;</span> <span class="n">i</span> <span class="o">!=</span> <span class="n">e</span><span class="p">;</span> <span class="o">++</span><span class="n">i</span><span class="p">)</span>
+      <span class="n">Visitor</span><span class="p">(</span><span class="o">&</span><span class="n">R</span><span class="o">-></span><span class="n">Roots</span><span class="p">[</span><span class="n">i</span><span class="p">],</span> <span class="n">R</span><span class="o">-></span><span class="n">Map</span><span class="o">-></span><span class="n">Meta</span><span class="p">[</span><span class="n">i</span><span class="p">]);</span>
+
+    <span class="c1">// For roots [NumMeta, NumRoots), the metadata pointer is null.</span>
+    <span class="k">for</span> <span class="p">(</span><span class="kt">unsigned</span> <span class="n">e</span> <span class="o">=</span> <span class="n">R</span><span class="o">-></span><span class="n">Map</span><span class="o">-></span><span class="n">NumRoots</span><span class="p">;</span> <span class="n">i</span> <span class="o">!=</span> <span class="n">e</span><span class="p">;</span> <span class="o">++</span><span class="n">i</span><span class="p">)</span>
+      <span class="n">Visitor</span><span class="p">(</span><span class="o">&</span><span class="n">R</span><span class="o">-></span><span class="n">Roots</span><span class="p">[</span><span class="n">i</span><span class="p">],</span> <span class="nb">NULL</span><span class="p">);</span>
+  <span class="p">}</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="about-the-shadow-stack">
+<h3><a class="toc-backref" href="#id6">About the shadow stack</a><a class="headerlink" href="#about-the-shadow-stack" title="Permalink to this headline">¶</a></h3>
+<p>Unlike many GC algorithms which rely on a cooperative code generator to compile
+stack maps, this algorithm carefully maintains a linked list of stack roots
+[<a class="reference internal" href="#henderson02"><em>Henderson2002</em></a>].  This so-called “shadow stack” mirrors the
+machine stack.  Maintaining this data structure is slower than using a stack map
+compiled into the executable as constant data, but has a significant portability
+advantage because it requires no special support from the target code generator,
+and does not require tricky platform-specific code to crawl the machine stack.</p>
+<p>The tradeoff for this simplicity and portability is:</p>
+<ul class="simple">
+<li>High overhead per function call.</li>
+<li>Not thread-safe.</li>
+</ul>
+<p>Still, it’s an easy way to get started.  After your compiler and runtime are up
+and running, writing a <a class="reference internal" href="#plugin"><em>plugin</em></a> will allow you to take advantage
+of <a class="reference internal" href="#collector-algos"><em>more advanced GC features</em></a> of LLVM in order to
+improve performance.</p>
+</div>
+</div>
+<div class="section" id="ir-features">
+<span id="gc-intrinsics"></span><h2><a class="toc-backref" href="#id7">IR features</a><a class="headerlink" href="#ir-features" title="Permalink to this headline">¶</a></h2>
+<p>This section describes the garbage collection facilities provided by the
+<a class="reference internal" href="LangRef.html"><em>LLVM intermediate representation</em></a>.  The exact behavior of these
+IR features is specified by the binary interface implemented by a <a class="reference internal" href="#plugin"><em>code
+generation plugin</em></a>, not by this document.</p>
+<p>These facilities are limited to those strictly necessary; they are not intended
+to be a complete interface to any garbage collector.  A program will need to
+interface with the GC library using the facilities provided by that program.</p>
+<div class="section" id="specifying-gc-code-generation-gc">
+<h3><a class="toc-backref" href="#id8">Specifying GC code generation: <tt class="docutils literal"><span class="pre">gc</span> <span class="pre">"..."</span></tt></a><a class="headerlink" href="#specifying-gc-code-generation-gc" title="Permalink to this headline">¶</a></h3>
+<div class="highlight-llvm"><pre>define ty @name(...) gc "name" { ...</pre>
+</div>
+<p>The <tt class="docutils literal"><span class="pre">gc</span></tt> function attribute is used to specify the desired GC style to the
+compiler.  Its programmatic equivalent is the <tt class="docutils literal"><span class="pre">setGC</span></tt> method of <tt class="docutils literal"><span class="pre">Function</span></tt>.</p>
+<p>Setting <tt class="docutils literal"><span class="pre">gc</span> <span class="pre">"name"</span></tt> on a function triggers a search for a matching code
+generation plugin “<em>name</em>”; it is that plugin which defines the exact nature of
+the code generated to support GC.  If none is found, the compiler will raise an
+error.</p>
+<p>Specifying the GC style on a per-function basis allows LLVM to link together
+programs that use different garbage collection algorithms (or none at all).</p>
+</div>
+<div class="section" id="gcroot">
+<span id="identifying-gc-roots-on-the-stack-llvm-gcroot"></span><h3><a class="toc-backref" href="#id9">Identifying GC roots on the stack: <tt class="docutils literal"><span class="pre">llvm.gcroot</span></tt></a><a class="headerlink" href="#gcroot" title="Permalink to this headline">¶</a></h3>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="kt">void</span> <span class="vg">@llvm.gcroot</span><span class="p">(</span><span class="k">i8</span><span class="p">**</span> <span class="nv">%ptrloc</span><span class="p">,</span> <span class="k">i8</span><span class="p">*</span> <span class="nv">%metadata</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>The <tt class="docutils literal"><span class="pre">llvm.gcroot</span></tt> intrinsic is used to inform LLVM that a stack variable
+references an object on the heap and is to be tracked for garbage collection.
+The exact impact on generated code is specified by a <a class="reference internal" href="#plugin"><em>compiler plugin</em></a>.  All calls to <tt class="docutils literal"><span class="pre">llvm.gcroot</span></tt> <strong>must</strong> reside inside the first basic
+block.</p>
+<p>A compiler which uses mem2reg to raise imperative code using <tt class="docutils literal"><span class="pre">alloca</span></tt> into SSA
+form need only add a call to <tt class="docutils literal"><span class="pre">@llvm.gcroot</span></tt> for those variables which a
+pointers into the GC heap.</p>
+<p>It is also important to mark intermediate values with <tt class="docutils literal"><span class="pre">llvm.gcroot</span></tt>.  For
+example, consider <tt class="docutils literal"><span class="pre">h(f(),</span> <span class="pre">g())</span></tt>.  Beware leaking the result of <tt class="docutils literal"><span class="pre">f()</span></tt> in the
+case that <tt class="docutils literal"><span class="pre">g()</span></tt> triggers a collection.  Note, that stack variables must be
+initialized and marked with <tt class="docutils literal"><span class="pre">llvm.gcroot</span></tt> in function’s prologue.</p>
+<p>The first argument <strong>must</strong> be a value referring to an alloca instruction or a
+bitcast of an alloca.  The second contains a pointer to metadata that should be
+associated with the pointer, and <strong>must</strong> be a constant or global value
+address.  If your target collector uses tags, use a null pointer for metadata.</p>
+<p>The <tt class="docutils literal"><span class="pre">%metadata</span></tt> argument can be used to avoid requiring heap objects to have
+‘isa’ pointers or tag bits. [<a class="reference internal" href="#appel89">Appel89</a>, <a class="reference internal" href="#goldberg91">Goldberg91</a>, <a class="reference internal" href="#tolmach94">Tolmach94</a>] If specified,
+its value will be tracked along with the location of the pointer in the stack
+frame.</p>
+<p>Consider the following fragment of Java code:</p>
+<div class="highlight-java"><div class="highlight"><pre><span class="o">{</span>
+  <span class="n">Object</span> <span class="n">X</span><span class="o">;</span>   <span class="c1">// A null-initialized reference to an object</span>
+  <span class="o">...</span>
+<span class="o">}</span>
+</pre></div>
+</div>
+<p>This block (which may be located in the middle of a function or in a loop nest),
+could be compiled to this LLVM code:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="nl">Entry:</span>
+   <span class="c">;; In the entry block for the function, allocate the</span>
+   <span class="c">;; stack space for X, which is an LLVM pointer.</span>
+   <span class="nv">%X</span> <span class="p">=</span> <span class="k">alloca</span> <span class="nv">%Object</span><span class="p">*</span>
+
+   <span class="c">;; Tell LLVM that the stack space is a stack root.</span>
+   <span class="c">;; Java has type-tags on objects, so we pass null as metadata.</span>
+   <span class="nv">%tmp</span> <span class="p">=</span> <span class="k">bitcast</span> <span class="nv">%Object</span><span class="p">**</span> <span class="nv">%X</span> <span class="k">to</span> <span class="k">i8</span><span class="p">**</span>
+   <span class="k">call</span> <span class="kt">void</span> <span class="vg">@llvm.gcroot</span><span class="p">(</span><span class="k">i8</span><span class="p">**</span> <span class="nv">%tmp</span><span class="p">,</span> <span class="k">i8</span><span class="p">*</span> <span class="k">null</span><span class="p">)</span>
+   <span class="p">...</span>
+
+   <span class="c">;; "CodeBlock" is the block corresponding to the start</span>
+   <span class="c">;;  of the scope above.</span>
+<span class="nl">CodeBlock:</span>
+   <span class="c">;; Java null-initializes pointers.</span>
+   <span class="k">store</span> <span class="nv">%Object</span><span class="p">*</span> <span class="k">null</span><span class="p">,</span> <span class="nv">%Object</span><span class="p">**</span> <span class="nv">%X</span>
+
+   <span class="p">...</span>
+
+   <span class="c">;; As the pointer goes out of scope, store a null value into</span>
+   <span class="c">;; it, to indicate that the value is no longer live.</span>
+   <span class="k">store</span> <span class="nv">%Object</span><span class="p">*</span> <span class="k">null</span><span class="p">,</span> <span class="nv">%Object</span><span class="p">**</span> <span class="nv">%X</span>
+   <span class="p">...</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="reading-and-writing-references-in-the-heap">
+<h3><a class="toc-backref" href="#id10">Reading and writing references in the heap</a><a class="headerlink" href="#reading-and-writing-references-in-the-heap" title="Permalink to this headline">¶</a></h3>
+<p>Some collectors need to be informed when the mutator (the program that needs
+garbage collection) either reads a pointer from or writes a pointer to a field
+of a heap object.  The code fragments inserted at these points are called <em>read
+barriers</em> and <em>write barriers</em>, respectively.  The amount of code that needs to
+be executed is usually quite small and not on the critical path of any
+computation, so the overall performance impact of the barrier is tolerable.</p>
+<p>Barriers often require access to the <em>object pointer</em> rather than the <em>derived
+pointer</em> (which is a pointer to the field within the object).  Accordingly,
+these intrinsics take both pointers as separate arguments for completeness.  In
+this snippet, <tt class="docutils literal"><span class="pre">%object</span></tt> is the object pointer, and <tt class="docutils literal"><span class="pre">%derived</span></tt> is the derived
+pointer:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="c">;; An array type.</span>
+<span class="nv">%class.Array</span> <span class="p">=</span> <span class="k">type</span> <span class="p">{</span> <span class="nv">%class.Object</span><span class="p">,</span> <span class="k">i32</span><span class="p">,</span> <span class="p">[</span><span class="m">0</span> <span class="k">x</span> <span class="nv">%class.Object</span><span class="p">*]</span> <span class="p">}</span>
+<span class="p">...</span>
+
+<span class="c">;; Load the object pointer from a gcroot.</span>
+<span class="nv">%object</span> <span class="p">=</span> <span class="k">load</span> <span class="nv">%class.Array</span><span class="p">**</span> <span class="nv">%object_addr</span>
+
+<span class="c">;; Compute the derived pointer.</span>
+<span class="nv">%derived</span> <span class="p">=</span> <span class="k">getelementptr</span> <span class="nv">%object</span><span class="p">,</span> <span class="k">i32</span> <span class="m">0</span><span class="p">,</span> <span class="k">i32</span> <span class="m">2</span><span class="p">,</span> <span class="k">i32</span> <span class="nv">%n</span>
+</pre></div>
+</div>
+<p>LLVM does not enforce this relationship between the object and derived pointer
+(although a <a class="reference internal" href="#plugin"><em>plugin</em></a> might).  However, it would be an unusual
+collector that violated it.</p>
+<p>The use of these intrinsics is naturally optional if the target GC does require
+the corresponding barrier.  Such a GC plugin will replace the intrinsic calls
+with the corresponding <tt class="docutils literal"><span class="pre">load</span></tt> or <tt class="docutils literal"><span class="pre">store</span></tt> instruction if they are used.</p>
+<div class="section" id="write-barrier-llvm-gcwrite">
+<h4><a class="toc-backref" href="#id11">Write barrier: <tt class="docutils literal"><span class="pre">llvm.gcwrite</span></tt></a><a class="headerlink" href="#write-barrier-llvm-gcwrite" title="Permalink to this headline">¶</a></h4>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="kt">void</span> <span class="vg">@llvm.gcwrite</span><span class="p">(</span><span class="k">i8</span><span class="p">*</span> <span class="nv">%value</span><span class="p">,</span> <span class="k">i8</span><span class="p">*</span> <span class="nv">%object</span><span class="p">,</span> <span class="k">i8</span><span class="p">**</span> <span class="nv">%derived</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>For write barriers, LLVM provides the <tt class="docutils literal"><span class="pre">llvm.gcwrite</span></tt> intrinsic function.  It
+has exactly the same semantics as a non-volatile <tt class="docutils literal"><span class="pre">store</span></tt> to the derived
+pointer (the third argument).  The exact code generated is specified by a
+compiler <a class="reference internal" href="#plugin"><em>plugin</em></a>.</p>
+<p>Many important algorithms require write barriers, including generational and
+concurrent collectors.  Additionally, write barriers could be used to implement
+reference counting.</p>
+</div>
+<div class="section" id="read-barrier-llvm-gcread">
+<h4><a class="toc-backref" href="#id12">Read barrier: <tt class="docutils literal"><span class="pre">llvm.gcread</span></tt></a><a class="headerlink" href="#read-barrier-llvm-gcread" title="Permalink to this headline">¶</a></h4>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="k">i8</span><span class="p">*</span> <span class="vg">@llvm.gcread</span><span class="p">(</span><span class="k">i8</span><span class="p">*</span> <span class="nv">%object</span><span class="p">,</span> <span class="k">i8</span><span class="p">**</span> <span class="nv">%derived</span><span class="p">)</span>
+</pre></div>
+</div>
+<p>For read barriers, LLVM provides the <tt class="docutils literal"><span class="pre">llvm.gcread</span></tt> intrinsic function.  It has
+exactly the same semantics as a non-volatile <tt class="docutils literal"><span class="pre">load</span></tt> from the derived pointer
+(the second argument).  The exact code generated is specified by a
+<a class="reference internal" href="#plugin"><em>compiler plugin</em></a>.</p>
+<p>Read barriers are needed by fewer algorithms than write barriers, and may have a
+greater performance impact since pointer reads are more frequent than writes.</p>
+</div>
+</div>
+</div>
+<div class="section" id="implementing-a-collector-plugin">
+<span id="plugin"></span><h2><a class="toc-backref" href="#id13">Implementing a collector plugin</a><a class="headerlink" href="#implementing-a-collector-plugin" title="Permalink to this headline">¶</a></h2>
+<p>User code specifies which GC code generation to use with the <tt class="docutils literal"><span class="pre">gc</span></tt> function
+attribute or, equivalently, with the <tt class="docutils literal"><span class="pre">setGC</span></tt> method of <tt class="docutils literal"><span class="pre">Function</span></tt>.</p>
+<p>To implement a GC plugin, it is necessary to subclass <tt class="docutils literal"><span class="pre">llvm::GCStrategy</span></tt>,
+which can be accomplished in a few lines of boilerplate code.  LLVM’s
+infrastructure provides access to several important algorithms.  For an
+uncontroversial collector, all that remains may be to compile LLVM’s computed
+stack map to assembly code (using the binary representation expected by the
+runtime library).  This can be accomplished in about 100 lines of code.</p>
+<p>This is not the appropriate place to implement a garbage collected heap or a
+garbage collector itself.  That code should exist in the language’s runtime
+library.  The compiler plugin is responsible for generating code which conforms
+to the binary interface defined by library, most essentially the <a class="reference internal" href="#stack-map"><em>stack map</em></a>.</p>
+<p>To subclass <tt class="docutils literal"><span class="pre">llvm::GCStrategy</span></tt> and register it with the compiler:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="c1">// lib/MyGC/MyGC.cpp - Example LLVM GC plugin</span>
+
+<span class="cp">#include "llvm/CodeGen/GCStrategy.h"</span>
+<span class="cp">#include "llvm/CodeGen/GCMetadata.h"</span>
+<span class="cp">#include "llvm/Support/Compiler.h"</span>
+
+<span class="k">using</span> <span class="k">namespace</span> <span class="n">llvm</span><span class="p">;</span>
+
+<span class="k">namespace</span> <span class="p">{</span>
+  <span class="k">class</span> <span class="nc">LLVM_LIBRARY_VISIBILITY</span> <span class="n">MyGC</span> <span class="o">:</span> <span class="k">public</span> <span class="n">GCStrategy</span> <span class="p">{</span>
+  <span class="nl">public:</span>
+    <span class="n">MyGC</span><span class="p">()</span> <span class="p">{}</span>
+  <span class="p">};</span>
+
+  <span class="n">GCRegistry</span><span class="o">::</span><span class="n">Add</span><span class="o"><</span><span class="n">MyGC</span><span class="o">></span>
+  <span class="n">X</span><span class="p">(</span><span class="s">"mygc"</span><span class="p">,</span> <span class="s">"My bespoke garbage collector."</span><span class="p">);</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>This boilerplate collector does nothing.  More specifically:</p>
+<ul class="simple">
+<li><tt class="docutils literal"><span class="pre">llvm.gcread</span></tt> calls are replaced with the corresponding <tt class="docutils literal"><span class="pre">load</span></tt>
+instruction.</li>
+<li><tt class="docutils literal"><span class="pre">llvm.gcwrite</span></tt> calls are replaced with the corresponding <tt class="docutils literal"><span class="pre">store</span></tt>
+instruction.</li>
+<li>No safe points are added to the code.</li>
+<li>The stack map is not compiled into the executable.</li>
+</ul>
+<p>Using the LLVM makefiles, this code
+can be compiled as a plugin using a simple makefile:</p>
+<div class="highlight-make"><div class="highlight"><pre><span class="c"># lib/MyGC/Makefile</span>
+
+<span class="nv">LEVEL</span> <span class="o">:=</span> ../..
+<span class="nv">LIBRARYNAME</span> <span class="o">=</span> MyGC
+<span class="nv">LOADABLE_MODULE</span> <span class="o">=</span> 1
+
+<span class="cp">include $(LEVEL)/Makefile.common</span>
+</pre></div>
+</div>
+<p>Once the plugin is compiled, code using it may be compiled using <tt class="docutils literal"><span class="pre">llc</span>
+<span class="pre">-load=MyGC.so</span></tt> (though MyGC.so may have some other platform-specific
+extension):</p>
+<div class="highlight-python"><pre>$ cat sample.ll
+define void @f() gc "mygc" {
+entry:
+  ret void
+}
+$ llvm-as < sample.ll | llc -load=MyGC.so</pre>
+</div>
+<p>It is also possible to statically link the collector plugin into tools, such as
+a language-specific compiler front-end.</p>
+<div class="section" id="overview-of-available-features">
+<span id="collector-algos"></span><h3><a class="toc-backref" href="#id14">Overview of available features</a><a class="headerlink" href="#overview-of-available-features" title="Permalink to this headline">¶</a></h3>
+<p><tt class="docutils literal"><span class="pre">GCStrategy</span></tt> provides a range of features through which a plugin may do useful
+work.  Some of these are callbacks, some are algorithms that can be enabled,
+disabled, or customized.  This matrix summarizes the supported (and planned)
+features and correlates them with the collection techniques which typically
+require them.</p>
+<table border="1" class="docutils">
+<colgroup>
+<col width="14%" />
+<col width="7%" />
+<col width="9%" />
+<col width="11%" />
+<col width="8%" />
+<col width="10%" />
+<col width="15%" />
+<col width="11%" />
+<col width="14%" />
+</colgroup>
+<thead valign="bottom">
+<tr class="row-odd"><th class="head">Algorithm</th>
+<th class="head">Done</th>
+<th class="head">Shadow
+stack</th>
+<th class="head">refcount</th>
+<th class="head">mark-
+sweep</th>
+<th class="head">copying</th>
+<th class="head">incremental</th>
+<th class="head">threaded</th>
+<th class="head">concurrent</th>
+</tr>
+</thead>
+<tbody valign="top">
+<tr class="row-even"><td>stack map</td>
+<td>✔</td>
+<td> </td>
+<td> </td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+</tr>
+<tr class="row-odd"><td>initialize
+roots</td>
+<td>✔</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+</tr>
+<tr class="row-even"><td>derived
+pointers</td>
+<td>NO</td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td><strong>N</strong>*</td>
+<td><strong>N</strong>*</td>
+</tr>
+<tr class="row-odd"><td><strong>custom
+lowering</strong></td>
+<td>✔</td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+</tr>
+<tr class="row-even"><td><em>gcroot</em></td>
+<td>✔</td>
+<td>✘</td>
+<td>✘</td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+</tr>
+<tr class="row-odd"><td><em>gcwrite</em></td>
+<td>✔</td>
+<td> </td>
+<td>✘</td>
+<td> </td>
+<td> </td>
+<td>✘</td>
+<td> </td>
+<td>✘</td>
+</tr>
+<tr class="row-even"><td><em>gcread</em></td>
+<td>✔</td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td>✘</td>
+</tr>
+<tr class="row-odd"><td><strong>safe
+points</strong></td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+</tr>
+<tr class="row-even"><td><em>in
+calls</em></td>
+<td>✔</td>
+<td> </td>
+<td> </td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+</tr>
+<tr class="row-odd"><td><em>before
+calls</em></td>
+<td>✔</td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td>✘</td>
+<td>✘</td>
+</tr>
+<tr class="row-even"><td><em>for
+loops</em></td>
+<td>NO</td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td><strong>N</strong></td>
+<td><strong>N</strong></td>
+</tr>
+<tr class="row-odd"><td><em>before
+escape</em></td>
+<td>✔</td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td>✘</td>
+<td>✘</td>
+</tr>
+<tr class="row-even"><td>emit code
+at safe
+points</td>
+<td>NO</td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td><strong>N</strong></td>
+<td><strong>N</strong></td>
+</tr>
+<tr class="row-odd"><td><strong>output</strong></td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+<td> </td>
+</tr>
+<tr class="row-even"><td><em>assembly</em></td>
+<td>✔</td>
+<td> </td>
+<td> </td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+<td>✘</td>
+</tr>
+<tr class="row-odd"><td><em>JIT</em></td>
+<td>NO</td>
+<td> </td>
+<td> </td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+</tr>
+<tr class="row-even"><td><em>obj</em></td>
+<td>NO</td>
+<td> </td>
+<td> </td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+</tr>
+<tr class="row-odd"><td>live
+analysis</td>
+<td>NO</td>
+<td> </td>
+<td> </td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+</tr>
+<tr class="row-even"><td>register
+map</td>
+<td>NO</td>
+<td> </td>
+<td> </td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+<td><strong>?</strong></td>
+</tr>
+<tr class="row-odd"><td colspan="9">* Derived pointers only pose a hasard to copying collections.</td>
+</tr>
+<tr class="row-even"><td colspan="9"><strong>?</strong> denotes a feature which could be utilized if available.</td>
+</tr>
+</tbody>
+</table>
+<p>To be clear, the collection techniques above are defined as:</p>
+<dl class="docutils">
+<dt>Shadow Stack</dt>
+<dd>The mutator carefully maintains a linked list of stack roots.</dd>
+<dt>Reference Counting</dt>
+<dd>The mutator maintains a reference count for each object and frees an object
+when its count falls to zero.</dd>
+<dt>Mark-Sweep</dt>
+<dd>When the heap is exhausted, the collector marks reachable objects starting
+from the roots, then deallocates unreachable objects in a sweep phase.</dd>
+<dt>Copying</dt>
+<dd>As reachability analysis proceeds, the collector copies objects from one heap
+area to another, compacting them in the process.  Copying collectors enable
+highly efficient “bump pointer” allocation and can improve locality of
+reference.</dd>
+<dt>Incremental</dt>
+<dd>(Including generational collectors.) Incremental collectors generally have all
+the properties of a copying collector (regardless of whether the mature heap
+is compacting), but bring the added complexity of requiring write barriers.</dd>
+<dt>Threaded</dt>
+<dd>Denotes a multithreaded mutator; the collector must still stop the mutator
+(“stop the world”) before beginning reachability analysis.  Stopping a
+multithreaded mutator is a complicated problem.  It generally requires highly
+platform-specific code in the runtime, and the production of carefully
+designed machine code at safe points.</dd>
+<dt>Concurrent</dt>
+<dd>In this technique, the mutator and the collector run concurrently, with the
+goal of eliminating pause times.  In a <em>cooperative</em> collector, the mutator
+further aids with collection should a pause occur, allowing collection to take
+advantage of multiprocessor hosts.  The “stop the world” problem of threaded
+collectors is generally still present to a limited extent.  Sophisticated
+marking algorithms are necessary.  Read barriers may be necessary.</dd>
+</dl>
+<p>As the matrix indicates, LLVM’s garbage collection infrastructure is already
+suitable for a wide variety of collectors, but does not currently extend to
+multithreaded programs.  This will be added in the future as there is
+interest.</p>
+</div>
+<div class="section" id="computing-stack-maps">
+<span id="stack-map"></span><h3><a class="toc-backref" href="#id15">Computing stack maps</a><a class="headerlink" href="#computing-stack-maps" title="Permalink to this headline">¶</a></h3>
+<p>LLVM automatically computes a stack map.  One of the most important features
+of a <tt class="docutils literal"><span class="pre">GCStrategy</span></tt> is to compile this information into the executable in
+the binary representation expected by the runtime library.</p>
+<p>The stack map consists of the location and identity of each GC root in the
+each function in the module.  For each root:</p>
+<ul class="simple">
+<li><tt class="docutils literal"><span class="pre">RootNum</span></tt>: The index of the root.</li>
+<li><tt class="docutils literal"><span class="pre">StackOffset</span></tt>: The offset of the object relative to the frame pointer.</li>
+<li><tt class="docutils literal"><span class="pre">RootMetadata</span></tt>: The value passed as the <tt class="docutils literal"><span class="pre">%metadata</span></tt> parameter to the
+<tt class="docutils literal"><span class="pre">@llvm.gcroot</span></tt> intrinsic.</li>
+</ul>
+<p>Also, for the function as a whole:</p>
+<ul>
+<li><dl class="first docutils">
+<dt><tt class="docutils literal"><span class="pre">getFrameSize()</span></tt>: The overall size of the function’s initial stack frame,</dt>
+<dd><p class="first last">not accounting for any dynamic allocation.</p>
+</dd>
+</dl>
+</li>
+<li><p class="first"><tt class="docutils literal"><span class="pre">roots_size()</span></tt>: The count of roots in the function.</p>
+</li>
+</ul>
+<p>To access the stack map, use <tt class="docutils literal"><span class="pre">GCFunctionMetadata::roots_begin()</span></tt> and
+-<tt class="docutils literal"><span class="pre">end()</span></tt> from the <a class="reference internal" href="#assembly"><em>GCMetadataPrinter</em></a>:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">for</span> <span class="p">(</span><span class="n">iterator</span> <span class="n">I</span> <span class="o">=</span> <span class="n">begin</span><span class="p">(),</span> <span class="n">E</span> <span class="o">=</span> <span class="n">end</span><span class="p">();</span> <span class="n">I</span> <span class="o">!=</span> <span class="n">E</span><span class="p">;</span> <span class="o">++</span><span class="n">I</span><span class="p">)</span> <span class="p">{</span>
+  <span class="n">GCFunctionInfo</span> <span class="o">*</span><span class="n">FI</span> <span class="o">=</span> <span class="o">*</span><span class="n">I</span><span class="p">;</span>
+  <span class="kt">unsigned</span> <span class="n">FrameSize</span> <span class="o">=</span> <span class="n">FI</span><span class="o">-></span><span class="n">getFrameSize</span><span class="p">();</span>
+  <span class="kt">size_t</span> <span class="n">RootCount</span> <span class="o">=</span> <span class="n">FI</span><span class="o">-></span><span class="n">roots_size</span><span class="p">();</span>
+
+  <span class="k">for</span> <span class="p">(</span><span class="n">GCFunctionInfo</span><span class="o">::</span><span class="n">roots_iterator</span> <span class="n">RI</span> <span class="o">=</span> <span class="n">FI</span><span class="o">-></span><span class="n">roots_begin</span><span class="p">(),</span>
+                                      <span class="n">RE</span> <span class="o">=</span> <span class="n">FI</span><span class="o">-></span><span class="n">roots_end</span><span class="p">();</span>
+                                      <span class="n">RI</span> <span class="o">!=</span> <span class="n">RE</span><span class="p">;</span> <span class="o">++</span><span class="n">RI</span><span class="p">)</span> <span class="p">{</span>
+    <span class="kt">int</span> <span class="n">RootNum</span> <span class="o">=</span> <span class="n">RI</span><span class="o">-></span><span class="n">Num</span><span class="p">;</span>
+    <span class="kt">int</span> <span class="n">RootStackOffset</span> <span class="o">=</span> <span class="n">RI</span><span class="o">-></span><span class="n">StackOffset</span><span class="p">;</span>
+    <span class="n">Constant</span> <span class="o">*</span><span class="n">RootMetadata</span> <span class="o">=</span> <span class="n">RI</span><span class="o">-></span><span class="n">Metadata</span><span class="p">;</span>
+  <span class="p">}</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>If the <tt class="docutils literal"><span class="pre">llvm.gcroot</span></tt> intrinsic is eliminated before code generation by a
+custom lowering pass, LLVM will compute an empty stack map.  This may be useful
+for collector plugins which implement reference counting or a shadow stack.</p>
+</div>
+<div class="section" id="initializing-roots-to-null-initroots">
+<span id="init-roots"></span><h3><a class="toc-backref" href="#id16">Initializing roots to null: <tt class="docutils literal"><span class="pre">InitRoots</span></tt></a><a class="headerlink" href="#initializing-roots-to-null-initroots" title="Permalink to this headline">¶</a></h3>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">MyGC</span><span class="o">::</span><span class="n">MyGC</span><span class="p">()</span> <span class="p">{</span>
+  <span class="n">InitRoots</span> <span class="o">=</span> <span class="nb">true</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>When set, LLVM will automatically initialize each root to <tt class="docutils literal"><span class="pre">null</span></tt> upon entry to
+the function.  This prevents the GC’s sweep phase from visiting uninitialized
+pointers, which will almost certainly cause it to crash.  This initialization
+occurs before custom lowering, so the two may be used together.</p>
+<p>Since LLVM does not yet compute liveness information, there is no means of
+distinguishing an uninitialized stack root from an initialized one.  Therefore,
+this feature should be used by all GC plugins.  It is enabled by default.</p>
+</div>
+<div class="section" id="custom-lowering-of-intrinsics-customroots-customreadbarriers-and-customwritebarriers">
+<h3><a class="toc-backref" href="#id17">Custom lowering of intrinsics: <tt class="docutils literal"><span class="pre">CustomRoots</span></tt>, <tt class="docutils literal"><span class="pre">CustomReadBarriers</span></tt>, and <tt class="docutils literal"><span class="pre">CustomWriteBarriers</span></tt></a><a class="headerlink" href="#custom-lowering-of-intrinsics-customroots-customreadbarriers-and-customwritebarriers" title="Permalink to this headline">¶</a></h3>
+<p>For GCs which use barriers or unusual treatment of stack roots, these flags
+allow the collector to perform arbitrary transformations of the LLVM IR:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">class</span> <span class="nc">MyGC</span> <span class="o">:</span> <span class="k">public</span> <span class="n">GCStrategy</span> <span class="p">{</span>
+<span class="nl">public:</span>
+  <span class="n">MyGC</span><span class="p">()</span> <span class="p">{</span>
+    <span class="n">CustomRoots</span> <span class="o">=</span> <span class="nb">true</span><span class="p">;</span>
+    <span class="n">CustomReadBarriers</span> <span class="o">=</span> <span class="nb">true</span><span class="p">;</span>
+    <span class="n">CustomWriteBarriers</span> <span class="o">=</span> <span class="nb">true</span><span class="p">;</span>
+  <span class="p">}</span>
+
+  <span class="k">virtual</span> <span class="kt">bool</span> <span class="n">initializeCustomLowering</span><span class="p">(</span><span class="n">Module</span> <span class="o">&</span><span class="n">M</span><span class="p">);</span>
+  <span class="k">virtual</span> <span class="kt">bool</span> <span class="nf">performCustomLowering</span><span class="p">(</span><span class="n">Function</span> <span class="o">&</span><span class="n">F</span><span class="p">);</span>
+<span class="p">};</span>
+</pre></div>
+</div>
+<p>If any of these flags are set, then LLVM suppresses its default lowering for the
+corresponding intrinsics and instead calls <tt class="docutils literal"><span class="pre">performCustomLowering</span></tt>.</p>
+<p>LLVM’s default action for each intrinsic is as follows:</p>
+<ul class="simple">
+<li><tt class="docutils literal"><span class="pre">llvm.gcroot</span></tt>: Leave it alone.  The code generator must see it or the stack
+map will not be computed.</li>
+<li><tt class="docutils literal"><span class="pre">llvm.gcread</span></tt>: Substitute a <tt class="docutils literal"><span class="pre">load</span></tt> instruction.</li>
+<li><tt class="docutils literal"><span class="pre">llvm.gcwrite</span></tt>: Substitute a <tt class="docutils literal"><span class="pre">store</span></tt> instruction.</li>
+</ul>
+<p>If <tt class="docutils literal"><span class="pre">CustomReadBarriers</span></tt> or <tt class="docutils literal"><span class="pre">CustomWriteBarriers</span></tt> are specified, then
+<tt class="docutils literal"><span class="pre">performCustomLowering</span></tt> <strong>must</strong> eliminate the corresponding barriers.</p>
+<p><tt class="docutils literal"><span class="pre">performCustomLowering</span></tt> must comply with the same restrictions as
+<a class="reference internal" href="WritingAnLLVMPass.html#writing-an-llvm-pass-runonfunction"><em>FunctionPass::runOnFunction</em></a>
+Likewise, <tt class="docutils literal"><span class="pre">initializeCustomLowering</span></tt> has the same semantics as
+<a class="reference internal" href="WritingAnLLVMPass.html#writing-an-llvm-pass-doinitialization-mod"><em>Pass::doInitialization(Module&)</em></a></p>
+<p>The following can be used as a template:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#include "llvm/IR/Module.h"</span>
+<span class="cp">#include "llvm/IR/IntrinsicInst.h"</span>
+
+<span class="kt">bool</span> <span class="n">MyGC</span><span class="o">::</span><span class="n">initializeCustomLowering</span><span class="p">(</span><span class="n">Module</span> <span class="o">&</span><span class="n">M</span><span class="p">)</span> <span class="p">{</span>
+  <span class="k">return</span> <span class="nb">false</span><span class="p">;</span>
+<span class="p">}</span>
+
+<span class="kt">bool</span> <span class="n">MyGC</span><span class="o">::</span><span class="n">performCustomLowering</span><span class="p">(</span><span class="n">Function</span> <span class="o">&</span><span class="n">F</span><span class="p">)</span> <span class="p">{</span>
+  <span class="kt">bool</span> <span class="n">MadeChange</span> <span class="o">=</span> <span class="nb">false</span><span class="p">;</span>
+
+  <span class="k">for</span> <span class="p">(</span><span class="n">Function</span><span class="o">::</span><span class="n">iterator</span> <span class="n">BB</span> <span class="o">=</span> <span class="n">F</span><span class="p">.</span><span class="n">begin</span><span class="p">(),</span> <span class="n">E</span> <span class="o">=</span> <span class="n">F</span><span class="p">.</span><span class="n">end</span><span class="p">();</span> <span class="n">BB</span> <span class="o">!=</span> <span class="n">E</span><span class="p">;</span> <span class="o">++</span><span class="n">BB</span><span class="p">)</span>
+    <span class="k">for</span> <span class="p">(</span><span class="n">BasicBlock</span><span class="o">::</span><span class="n">iterator</span> <span class="n">II</span> <span class="o">=</span> <span class="n">BB</span><span class="o">-></span><span class="n">begin</span><span class="p">(),</span> <span class="n">E</span> <span class="o">=</span> <span class="n">BB</span><span class="o">-></span><span class="n">end</span><span class="p">();</span> <span class="n">II</span> <span class="o">!=</span> <span class="n">E</span><span class="p">;</span> <span class="p">)</span>
+      <span class="k">if</span> <span class="p">(</span><span class="n">IntrinsicInst</span> <span class="o">*</span><span class="n">CI</span> <span class="o">=</span> <span class="n">dyn_cast</span><span class="o"><</span><span class="n">IntrinsicInst</span><span class="o">></span><span class="p">(</span><span class="n">II</span><span class="o">++</span><span class="p">))</span>
+        <span class="k">if</span> <span class="p">(</span><span class="n">Function</span> <span class="o">*</span><span class="n">F</span> <span class="o">=</span> <span class="n">CI</span><span class="o">-></span><span class="n">getCalledFunction</span><span class="p">())</span>
+          <span class="k">switch</span> <span class="p">(</span><span class="n">F</span><span class="o">-></span><span class="n">getIntrinsicID</span><span class="p">())</span> <span class="p">{</span>
+          <span class="k">case</span> <span class="n">Intrinsic</span>:<span class="o">:</span><span class="n">gcwrite</span><span class="o">:</span>
+            <span class="c1">// Handle llvm.gcwrite.</span>
+            <span class="n">CI</span><span class="o">-></span><span class="n">eraseFromParent</span><span class="p">();</span>
+            <span class="n">MadeChange</span> <span class="o">=</span> <span class="nb">true</span><span class="p">;</span>
+            <span class="k">break</span><span class="p">;</span>
+          <span class="k">case</span> <span class="n">Intrinsic</span>:<span class="o">:</span><span class="n">gcread</span><span class="o">:</span>
+            <span class="c1">// Handle llvm.gcread.</span>
+            <span class="n">CI</span><span class="o">-></span><span class="n">eraseFromParent</span><span class="p">();</span>
+            <span class="n">MadeChange</span> <span class="o">=</span> <span class="nb">true</span><span class="p">;</span>
+            <span class="k">break</span><span class="p">;</span>
+          <span class="k">case</span> <span class="n">Intrinsic</span>:<span class="o">:</span><span class="n">gcroot</span><span class="o">:</span>
+            <span class="c1">// Handle llvm.gcroot.</span>
+            <span class="n">CI</span><span class="o">-></span><span class="n">eraseFromParent</span><span class="p">();</span>
+            <span class="n">MadeChange</span> <span class="o">=</span> <span class="nb">true</span><span class="p">;</span>
+            <span class="k">break</span><span class="p">;</span>
+          <span class="p">}</span>
+
+  <span class="k">return</span> <span class="n">MadeChange</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="generating-safe-points-neededsafepoints">
+<span id="safe-points"></span><h3><a class="toc-backref" href="#id18">Generating safe points: <tt class="docutils literal"><span class="pre">NeededSafePoints</span></tt></a><a class="headerlink" href="#generating-safe-points-neededsafepoints" title="Permalink to this headline">¶</a></h3>
+<p>LLVM can compute four kinds of safe points:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">namespace</span> <span class="n">GC</span> <span class="p">{</span>
+  <span class="c1">/// PointKind - The type of a collector-safe point.</span>
+  <span class="c1">///</span>
+  <span class="k">enum</span> <span class="n">PointKind</span> <span class="p">{</span>
+    <span class="n">Loop</span><span class="p">,</span>    <span class="c1">//< Instr is a loop (backwards branch).</span>
+    <span class="n">Return</span><span class="p">,</span>  <span class="c1">//< Instr is a return instruction.</span>
+    <span class="n">PreCall</span><span class="p">,</span> <span class="c1">//< Instr is a call instruction.</span>
+    <span class="n">PostCall</span> <span class="c1">//< Instr is the return address of a call.</span>
+  <span class="p">};</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>A collector can request any combination of the four by setting the
+<tt class="docutils literal"><span class="pre">NeededSafePoints</span></tt> mask:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">MyGC</span><span class="o">::</span><span class="n">MyGC</span><span class="p">()</span>  <span class="p">{</span>
+  <span class="n">NeededSafePoints</span> <span class="o">=</span> <span class="mi">1</span> <span class="o"><<</span> <span class="n">GC</span><span class="o">::</span><span class="n">Loop</span>
+                   <span class="o">|</span> <span class="mi">1</span> <span class="o"><<</span> <span class="n">GC</span><span class="o">::</span><span class="n">Return</span>
+                   <span class="o">|</span> <span class="mi">1</span> <span class="o"><<</span> <span class="n">GC</span><span class="o">::</span><span class="n">PreCall</span>
+                   <span class="o">|</span> <span class="mi">1</span> <span class="o"><<</span> <span class="n">GC</span><span class="o">::</span><span class="n">PostCall</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>It can then use the following routines to access safe points.</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">for</span> <span class="p">(</span><span class="n">iterator</span> <span class="n">I</span> <span class="o">=</span> <span class="n">begin</span><span class="p">(),</span> <span class="n">E</span> <span class="o">=</span> <span class="n">end</span><span class="p">();</span> <span class="n">I</span> <span class="o">!=</span> <span class="n">E</span><span class="p">;</span> <span class="o">++</span><span class="n">I</span><span class="p">)</span> <span class="p">{</span>
+  <span class="n">GCFunctionInfo</span> <span class="o">*</span><span class="n">MD</span> <span class="o">=</span> <span class="o">*</span><span class="n">I</span><span class="p">;</span>
+  <span class="kt">size_t</span> <span class="n">PointCount</span> <span class="o">=</span> <span class="n">MD</span><span class="o">-></span><span class="n">size</span><span class="p">();</span>
+
+  <span class="k">for</span> <span class="p">(</span><span class="n">GCFunctionInfo</span><span class="o">::</span><span class="n">iterator</span> <span class="n">PI</span> <span class="o">=</span> <span class="n">MD</span><span class="o">-></span><span class="n">begin</span><span class="p">(),</span>
+                                <span class="n">PE</span> <span class="o">=</span> <span class="n">MD</span><span class="o">-></span><span class="n">end</span><span class="p">();</span> <span class="n">PI</span> <span class="o">!=</span> <span class="n">PE</span><span class="p">;</span> <span class="o">++</span><span class="n">PI</span><span class="p">)</span> <span class="p">{</span>
+    <span class="n">GC</span><span class="o">::</span><span class="n">PointKind</span> <span class="n">PointKind</span> <span class="o">=</span> <span class="n">PI</span><span class="o">-></span><span class="n">Kind</span><span class="p">;</span>
+    <span class="kt">unsigned</span> <span class="n">PointNum</span> <span class="o">=</span> <span class="n">PI</span><span class="o">-></span><span class="n">Num</span><span class="p">;</span>
+  <span class="p">}</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>Almost every collector requires <tt class="docutils literal"><span class="pre">PostCall</span></tt> safe points, since these correspond
+to the moments when the function is suspended during a call to a subroutine.</p>
+<p>Threaded programs generally require <tt class="docutils literal"><span class="pre">Loop</span></tt> safe points to guarantee that the
+application will reach a safe point within a bounded amount of time, even if it
+is executing a long-running loop which contains no function calls.</p>
+<p>Threaded collectors may also require <tt class="docutils literal"><span class="pre">Return</span></tt> and <tt class="docutils literal"><span class="pre">PreCall</span></tt> safe points to
+implement “stop the world” techniques using self-modifying code, where it is
+important that the program not exit the function without reaching a safe point
+(because only the topmost function has been patched).</p>
+</div>
+<div class="section" id="emitting-assembly-code-gcmetadataprinter">
+<span id="assembly"></span><h3><a class="toc-backref" href="#id19">Emitting assembly code: <tt class="docutils literal"><span class="pre">GCMetadataPrinter</span></tt></a><a class="headerlink" href="#emitting-assembly-code-gcmetadataprinter" title="Permalink to this headline">¶</a></h3>
+<p>LLVM allows a plugin to print arbitrary assembly code before and after the rest
+of a module’s assembly code.  At the end of the module, the GC can compile the
+LLVM stack map into assembly code. (At the beginning, this information is not
+yet computed.)</p>
+<p>Since AsmWriter and CodeGen are separate components of LLVM, a separate abstract
+base class and registry is provided for printing assembly code, the
+<tt class="docutils literal"><span class="pre">GCMetadaPrinter</span></tt> and <tt class="docutils literal"><span class="pre">GCMetadataPrinterRegistry</span></tt>.  The AsmWriter will look
+for such a subclass if the <tt class="docutils literal"><span class="pre">GCStrategy</span></tt> sets <tt class="docutils literal"><span class="pre">UsesMetadata</span></tt>:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">MyGC</span><span class="o">::</span><span class="n">MyGC</span><span class="p">()</span> <span class="p">{</span>
+  <span class="n">UsesMetadata</span> <span class="o">=</span> <span class="nb">true</span><span class="p">;</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>This separation allows JIT-only clients to be smaller.</p>
+<p>Note that LLVM does not currently have analogous APIs to support code generation
+in the JIT, nor using the object writers.</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="c1">// lib/MyGC/MyGCPrinter.cpp - Example LLVM GC printer</span>
+
+<span class="cp">#include "llvm/CodeGen/GCMetadataPrinter.h"</span>
+<span class="cp">#include "llvm/Support/Compiler.h"</span>
+
+<span class="k">using</span> <span class="k">namespace</span> <span class="n">llvm</span><span class="p">;</span>
+
+<span class="k">namespace</span> <span class="p">{</span>
+  <span class="k">class</span> <span class="nc">LLVM_LIBRARY_VISIBILITY</span> <span class="n">MyGCPrinter</span> <span class="o">:</span> <span class="k">public</span> <span class="n">GCMetadataPrinter</span> <span class="p">{</span>
+  <span class="nl">public:</span>
+    <span class="k">virtual</span> <span class="kt">void</span> <span class="n">beginAssembly</span><span class="p">(</span><span class="n">AsmPrinter</span> <span class="o">&</span><span class="n">AP</span><span class="p">);</span>
+
+    <span class="k">virtual</span> <span class="kt">void</span> <span class="nf">finishAssembly</span><span class="p">(</span><span class="n">AsmPrinter</span> <span class="o">&</span><span class="n">AP</span><span class="p">);</span>
+  <span class="p">};</span>
+
+  <span class="n">GCMetadataPrinterRegistry</span><span class="o">::</span><span class="n">Add</span><span class="o"><</span><span class="n">MyGCPrinter</span><span class="o">></span>
+  <span class="n">X</span><span class="p">(</span><span class="s">"mygc"</span><span class="p">,</span> <span class="s">"My bespoke garbage collector."</span><span class="p">);</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>The collector should use <tt class="docutils literal"><span class="pre">AsmPrinter</span></tt> to print portable assembly code.  The
+collector itself contains the stack map for the entire module, and may access
+the <tt class="docutils literal"><span class="pre">GCFunctionInfo</span></tt> using its own <tt class="docutils literal"><span class="pre">begin()</span></tt> and <tt class="docutils literal"><span class="pre">end()</span></tt> methods.  Here’s
+a realistic example:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="cp">#include "llvm/CodeGen/AsmPrinter.h"</span>
+<span class="cp">#include "llvm/IR/Function.h"</span>
+<span class="cp">#include "llvm/IR/DataLayout.h"</span>
+<span class="cp">#include "llvm/Target/TargetAsmInfo.h"</span>
+<span class="cp">#include "llvm/Target/TargetMachine.h"</span>
+
+<span class="kt">void</span> <span class="n">MyGCPrinter</span><span class="o">::</span><span class="n">beginAssembly</span><span class="p">(</span><span class="n">AsmPrinter</span> <span class="o">&</span><span class="n">AP</span><span class="p">)</span> <span class="p">{</span>
+  <span class="c1">// Nothing to do.</span>
+<span class="p">}</span>
+
+<span class="kt">void</span> <span class="n">MyGCPrinter</span><span class="o">::</span><span class="n">finishAssembly</span><span class="p">(</span><span class="n">AsmPrinter</span> <span class="o">&</span><span class="n">AP</span><span class="p">)</span> <span class="p">{</span>
+  <span class="n">MCStreamer</span> <span class="o">&</span><span class="n">OS</span> <span class="o">=</span> <span class="n">AP</span><span class="p">.</span><span class="n">OutStreamer</span><span class="p">;</span>
+  <span class="kt">unsigned</span> <span class="n">IntPtrSize</span> <span class="o">=</span> <span class="n">AP</span><span class="p">.</span><span class="n">TM</span><span class="p">.</span><span class="n">getDataLayout</span><span class="p">()</span><span class="o">-></span><span class="n">getPointerSize</span><span class="p">();</span>
+
+  <span class="c1">// Put this in the data section.</span>
+  <span class="n">OS</span><span class="p">.</span><span class="n">SwitchSection</span><span class="p">(</span><span class="n">AP</span><span class="p">.</span><span class="n">getObjFileLowering</span><span class="p">().</span><span class="n">getDataSection</span><span class="p">());</span>
+
+  <span class="c1">// For each function...</span>
+  <span class="k">for</span> <span class="p">(</span><span class="n">iterator</span> <span class="n">FI</span> <span class="o">=</span> <span class="n">begin</span><span class="p">(),</span> <span class="n">FE</span> <span class="o">=</span> <span class="n">end</span><span class="p">();</span> <span class="n">FI</span> <span class="o">!=</span> <span class="n">FE</span><span class="p">;</span> <span class="o">++</span><span class="n">FI</span><span class="p">)</span> <span class="p">{</span>
+    <span class="n">GCFunctionInfo</span> <span class="o">&</span><span class="n">MD</span> <span class="o">=</span> <span class="o">**</span><span class="n">FI</span><span class="p">;</span>
+
+    <span class="c1">// A compact GC layout. Emit this data structure:</span>
+    <span class="c1">//</span>
+    <span class="c1">// struct {</span>
+    <span class="c1">//   int32_t PointCount;</span>
+    <span class="c1">//   void *SafePointAddress[PointCount];</span>
+    <span class="c1">//   int32_t StackFrameSize; // in words</span>
+    <span class="c1">//   int32_t StackArity;</span>
+    <span class="c1">//   int32_t LiveCount;</span>
+    <span class="c1">//   int32_t LiveOffsets[LiveCount];</span>
+    <span class="c1">// } __gcmap_<FUNCTIONNAME>;</span>
+
+    <span class="c1">// Align to address width.</span>
+    <span class="n">AP</span><span class="p">.</span><span class="n">EmitAlignment</span><span class="p">(</span><span class="n">IntPtrSize</span> <span class="o">==</span> <span class="mi">4</span> <span class="o">?</span> <span class="mi">2</span> <span class="o">:</span> <span class="mi">3</span><span class="p">);</span>
+
+    <span class="c1">// Emit PointCount.</span>
+    <span class="n">OS</span><span class="p">.</span><span class="n">AddComment</span><span class="p">(</span><span class="s">"safe point count"</span><span class="p">);</span>
+    <span class="n">AP</span><span class="p">.</span><span class="n">EmitInt32</span><span class="p">(</span><span class="n">MD</span><span class="p">.</span><span class="n">size</span><span class="p">());</span>
+
+    <span class="c1">// And each safe point...</span>
+    <span class="k">for</span> <span class="p">(</span><span class="n">GCFunctionInfo</span><span class="o">::</span><span class="n">iterator</span> <span class="n">PI</span> <span class="o">=</span> <span class="n">MD</span><span class="p">.</span><span class="n">begin</span><span class="p">(),</span>
+                                  <span class="n">PE</span> <span class="o">=</span> <span class="n">MD</span><span class="p">.</span><span class="n">end</span><span class="p">();</span> <span class="n">PI</span> <span class="o">!=</span> <span class="n">PE</span><span class="p">;</span> <span class="o">++</span><span class="n">PI</span><span class="p">)</span> <span class="p">{</span>
+      <span class="c1">// Emit the address of the safe point.</span>
+      <span class="n">OS</span><span class="p">.</span><span class="n">AddComment</span><span class="p">(</span><span class="s">"safe point address"</span><span class="p">);</span>
+      <span class="n">MCSymbol</span> <span class="o">*</span><span class="n">Label</span> <span class="o">=</span> <span class="n">PI</span><span class="o">-></span><span class="n">Label</span><span class="p">;</span>
+      <span class="n">AP</span><span class="p">.</span><span class="n">EmitLabelPlusOffset</span><span class="p">(</span><span class="n">Label</span><span class="cm">/*Hi*/</span><span class="p">,</span> <span class="mi">0</span><span class="cm">/*Offset*/</span><span class="p">,</span> <span class="mi">4</span><span class="cm">/*Size*/</span><span class="p">);</span>
+    <span class="p">}</span>
+
+    <span class="c1">// Stack information never change in safe points! Only print info from the</span>
+    <span class="c1">// first call-site.</span>
+    <span class="n">GCFunctionInfo</span><span class="o">::</span><span class="n">iterator</span> <span class="n">PI</span> <span class="o">=</span> <span class="n">MD</span><span class="p">.</span><span class="n">begin</span><span class="p">();</span>
+
+    <span class="c1">// Emit the stack frame size.</span>
+    <span class="n">OS</span><span class="p">.</span><span class="n">AddComment</span><span class="p">(</span><span class="s">"stack frame size (in words)"</span><span class="p">);</span>
+    <span class="n">AP</span><span class="p">.</span><span class="n">EmitInt32</span><span class="p">(</span><span class="n">MD</span><span class="p">.</span><span class="n">getFrameSize</span><span class="p">()</span> <span class="o">/</span> <span class="n">IntPtrSize</span><span class="p">);</span>
+
+    <span class="c1">// Emit stack arity, i.e. the number of stacked arguments.</span>
+    <span class="kt">unsigned</span> <span class="n">RegisteredArgs</span> <span class="o">=</span> <span class="n">IntPtrSize</span> <span class="o">==</span> <span class="mi">4</span> <span class="o">?</span> <span class="mi">5</span> <span class="o">:</span> <span class="mi">6</span><span class="p">;</span>
+    <span class="kt">unsigned</span> <span class="n">StackArity</span> <span class="o">=</span> <span class="n">MD</span><span class="p">.</span><span class="n">getFunction</span><span class="p">().</span><span class="n">arg_size</span><span class="p">()</span> <span class="o">></span> <span class="n">RegisteredArgs</span> <span class="o">?</span>
+                          <span class="n">MD</span><span class="p">.</span><span class="n">getFunction</span><span class="p">().</span><span class="n">arg_size</span><span class="p">()</span> <span class="o">-</span> <span class="n">RegisteredArgs</span> <span class="o">:</span> <span class="mi">0</span><span class="p">;</span>
+    <span class="n">OS</span><span class="p">.</span><span class="n">AddComment</span><span class="p">(</span><span class="s">"stack arity"</span><span class="p">);</span>
+    <span class="n">AP</span><span class="p">.</span><span class="n">EmitInt32</span><span class="p">(</span><span class="n">StackArity</span><span class="p">);</span>
+
+    <span class="c1">// Emit the number of live roots in the function.</span>
+    <span class="n">OS</span><span class="p">.</span><span class="n">AddComment</span><span class="p">(</span><span class="s">"live root count"</span><span class="p">);</span>
+    <span class="n">AP</span><span class="p">.</span><span class="n">EmitInt32</span><span class="p">(</span><span class="n">MD</span><span class="p">.</span><span class="n">live_size</span><span class="p">(</span><span class="n">PI</span><span class="p">));</span>
+
+    <span class="c1">// And for each live root...</span>
+    <span class="k">for</span> <span class="p">(</span><span class="n">GCFunctionInfo</span><span class="o">::</span><span class="n">live_iterator</span> <span class="n">LI</span> <span class="o">=</span> <span class="n">MD</span><span class="p">.</span><span class="n">live_begin</span><span class="p">(</span><span class="n">PI</span><span class="p">),</span>
+                                       <span class="n">LE</span> <span class="o">=</span> <span class="n">MD</span><span class="p">.</span><span class="n">live_end</span><span class="p">(</span><span class="n">PI</span><span class="p">);</span>
+                                       <span class="n">LI</span> <span class="o">!=</span> <span class="n">LE</span><span class="p">;</span> <span class="o">++</span><span class="n">LI</span><span class="p">)</span> <span class="p">{</span>
+      <span class="c1">// Emit live root's offset within the stack frame.</span>
+      <span class="n">OS</span><span class="p">.</span><span class="n">AddComment</span><span class="p">(</span><span class="s">"stack index (offset / wordsize)"</span><span class="p">);</span>
+      <span class="n">AP</span><span class="p">.</span><span class="n">EmitInt32</span><span class="p">(</span><span class="n">LI</span><span class="o">-></span><span class="n">StackOffset</span><span class="p">);</span>
+    <span class="p">}</span>
+  <span class="p">}</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+</div>
+</div>
+<div class="section" id="references">
+<h2><a class="toc-backref" href="#id20">References</a><a class="headerlink" href="#references" title="Permalink to this headline">¶</a></h2>
+<p id="appel89">[Appel89] Runtime Tags Aren’t Necessary. Andrew W. Appel. Lisp and Symbolic
+Computation 19(7):703-705, July 1989.</p>
+<p id="goldberg91">[Goldberg91] Tag-free garbage collection for strongly typed programming
+languages. Benjamin Goldberg. ACM SIGPLAN PLDI‘91.</p>
+<p id="tolmach94">[Tolmach94] Tag-free garbage collection using explicit type parameters. Andrew
+Tolmach. Proceedings of the 1994 ACM conference on LISP and functional
+programming.</p>
+<p id="henderson02">[Henderson2002] <a class="reference external" href="http://citeseer.ist.psu.edu/henderson02accurate.html">Accurate Garbage Collection in an Uncooperative Environment</a></p>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="WritingAnLLVMPass.html" title="Writing an LLVM Pass"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="HowToUseInstrMappings.html" title="How To Use Instruction Mappings"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/GetElementPtr.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/GetElementPtr.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/GetElementPtr.html (added)
+++ www-releases/trunk/3.5.1/docs/GetElementPtr.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,565 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>The Often Misunderstood GEP Instruction — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="MCJIT Design and Implementation" href="MCJITDesignAndImplementation.html" />
+    <link rel="prev" title="YAML I/O" href="YamlIO.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">
+      <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="MCJITDesignAndImplementation.html" title="MCJIT Design and Implementation"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="YamlIO.html" title="YAML I/O"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="the-often-misunderstood-gep-instruction">
+<h1>The Often Misunderstood GEP Instruction<a class="headerlink" href="#the-often-misunderstood-gep-instruction" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#introduction" id="id2">Introduction</a></li>
+<li><a class="reference internal" href="#address-computation" id="id3">Address Computation</a><ul>
+<li><a class="reference internal" href="#what-is-the-first-index-of-the-gep-instruction" id="id4">What is the first index of the GEP instruction?</a></li>
+<li><a class="reference internal" href="#why-is-the-extra-0-index-required" id="id5">Why is the extra 0 index required?</a></li>
+<li><a class="reference internal" href="#what-is-dereferenced-by-gep" id="id6">What is dereferenced by GEP?</a></li>
+<li><a class="reference internal" href="#why-don-t-gep-x-0-0-1-and-gep-x-1-alias" id="id7">Why don’t GEP x,0,0,1 and GEP x,1 alias?</a></li>
+<li><a class="reference internal" href="#why-do-gep-x-1-0-0-and-gep-x-1-alias" id="id8">Why do GEP x,1,0,0 and GEP x,1 alias?</a></li>
+<li><a class="reference internal" href="#can-gep-index-into-vector-elements" id="id9">Can GEP index into vector elements?</a></li>
+<li><a class="reference internal" href="#what-effect-do-address-spaces-have-on-geps" id="id10">What effect do address spaces have on GEPs?</a></li>
+<li><a class="reference internal" href="#how-is-gep-different-from-ptrtoint-arithmetic-and-inttoptr" id="id11">How is GEP different from <tt class="docutils literal"><span class="pre">ptrtoint</span></tt>, arithmetic, and <tt class="docutils literal"><span class="pre">inttoptr</span></tt>?</a></li>
+<li><a class="reference internal" href="#i-m-writing-a-backend-for-a-target-which-needs-custom-lowering-for-gep-how-do-i-do-this" id="id12">I’m writing a backend for a target which needs custom lowering for GEP. How do I do this?</a></li>
+<li><a class="reference internal" href="#how-does-vla-addressing-work-with-geps" id="id13">How does VLA addressing work with GEPs?</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#rules" id="id14">Rules</a><ul>
+<li><a class="reference internal" href="#what-happens-if-an-array-index-is-out-of-bounds" id="id15">What happens if an array index is out of bounds?</a></li>
+<li><a class="reference internal" href="#can-array-indices-be-negative" id="id16">Can array indices be negative?</a></li>
+<li><a class="reference internal" href="#can-i-compare-two-values-computed-with-geps" id="id17">Can I compare two values computed with GEPs?</a></li>
+<li><a class="reference internal" href="#can-i-do-gep-with-a-different-pointer-type-than-the-type-of-the-underlying-object" id="id18">Can I do GEP with a different pointer type than the type of the underlying object?</a></li>
+<li><a class="reference internal" href="#can-i-cast-an-object-s-address-to-integer-and-add-it-to-null" id="id19">Can I cast an object’s address to integer and add it to null?</a></li>
+<li><a class="reference internal" href="#can-i-compute-the-distance-between-two-objects-and-add-that-value-to-one-address-to-compute-the-other-address" id="id20">Can I compute the distance between two objects, and add that value to one address to compute the other address?</a></li>
+<li><a class="reference internal" href="#can-i-do-type-based-alias-analysis-on-llvm-ir" id="id21">Can I do type-based alias analysis on LLVM IR?</a></li>
+<li><a class="reference internal" href="#what-happens-if-a-gep-computation-overflows" id="id22">What happens if a GEP computation overflows?</a></li>
+<li><a class="reference internal" href="#how-can-i-tell-if-my-front-end-is-following-the-rules" id="id23">How can I tell if my front-end is following the rules?</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#rationale" id="id24">Rationale</a><ul>
+<li><a class="reference internal" href="#why-is-gep-designed-this-way" id="id25">Why is GEP designed this way?</a></li>
+<li><a class="reference internal" href="#why-do-struct-member-indices-always-use-i32" id="id26">Why do struct member indices always use <tt class="docutils literal"><span class="pre">i32</span></tt>?</a></li>
+<li><a class="reference internal" href="#what-s-an-uglygep" id="id27">What’s an uglygep?</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#summary" id="id28">Summary</a></li>
+</ul>
+</div>
+<div class="section" id="introduction">
+<h2><a class="toc-backref" href="#id2">Introduction</a><a class="headerlink" href="#introduction" title="Permalink to this headline">¶</a></h2>
+<p>This document seeks to dispel the mystery and confusion surrounding LLVM’s
+<a class="reference external" href="LangRef.html#i_getelementptr">GetElementPtr</a> (GEP) instruction.  Questions
+about the wily GEP instruction are probably the most frequently occurring
+questions once a developer gets down to coding with LLVM. Here we lay out the
+sources of confusion and show that the GEP instruction is really quite simple.</p>
+</div>
+<div class="section" id="address-computation">
+<h2><a class="toc-backref" href="#id3">Address Computation</a><a class="headerlink" href="#address-computation" title="Permalink to this headline">¶</a></h2>
+<p>When people are first confronted with the GEP instruction, they tend to relate
+it to known concepts from other programming paradigms, most notably C array
+indexing and field selection. GEP closely resembles C array indexing and field
+selection, however it is a little different and this leads to the following
+questions.</p>
+<div class="section" id="what-is-the-first-index-of-the-gep-instruction">
+<h3><a class="toc-backref" href="#id4">What is the first index of the GEP instruction?</a><a class="headerlink" href="#what-is-the-first-index-of-the-gep-instruction" title="Permalink to this headline">¶</a></h3>
+<p>Quick answer: The index stepping through the first operand.</p>
+<p>The confusion with the first index usually arises from thinking about the
+GetElementPtr instruction as if it was a C index operator. They aren’t the
+same. For example, when we write, in “C”:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">AType</span> <span class="o">*</span><span class="n">Foo</span><span class="p">;</span>
+<span class="p">...</span>
+<span class="n">X</span> <span class="o">=</span> <span class="o">&</span><span class="n">Foo</span><span class="o">-></span><span class="n">F</span><span class="p">;</span>
+</pre></div>
+</div>
+<p>it is natural to think that there is only one index, the selection of the field
+<tt class="docutils literal"><span class="pre">F</span></tt>.  However, in this example, <tt class="docutils literal"><span class="pre">Foo</span></tt> is a pointer. That pointer
+must be indexed explicitly in LLVM. C, on the other hand, indices through it
+transparently.  To arrive at the same address location as the C code, you would
+provide the GEP instruction with two index operands. The first operand indexes
+through the pointer; the second operand indexes the field <tt class="docutils literal"><span class="pre">F</span></tt> of the
+structure, just as if you wrote:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">X</span> <span class="o">=</span> <span class="o">&</span><span class="n">Foo</span><span class="p">[</span><span class="mi">0</span><span class="p">].</span><span class="n">F</span><span class="p">;</span>
+</pre></div>
+</div>
+<p>Sometimes this question gets rephrased as:</p>
+<blockquote id="gep-index-through-first-pointer">
+<div><em>Why is it okay to index through the first pointer, but subsequent pointers
+won’t be dereferenced?</em></div></blockquote>
+<p>The answer is simply because memory does not have to be accessed to perform the
+computation. The first operand to the GEP instruction must be a value of a
+pointer type. The value of the pointer is provided directly to the GEP
+instruction as an operand without any need for accessing memory. It must,
+therefore be indexed and requires an index operand. Consider this example:</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="k">struct</span> <span class="n">munger_struct</span> <span class="p">{</span>
+  <span class="kt">int</span> <span class="n">f1</span><span class="p">;</span>
+  <span class="kt">int</span> <span class="n">f2</span><span class="p">;</span>
+<span class="p">};</span>
+<span class="kt">void</span> <span class="nf">munge</span><span class="p">(</span><span class="k">struct</span> <span class="n">munger_struct</span> <span class="o">*</span><span class="n">P</span><span class="p">)</span> <span class="p">{</span>
+  <span class="n">P</span><span class="p">[</span><span class="mi">0</span><span class="p">].</span><span class="n">f1</span> <span class="o">=</span> <span class="n">P</span><span class="p">[</span><span class="mi">1</span><span class="p">].</span><span class="n">f1</span> <span class="o">+</span> <span class="n">P</span><span class="p">[</span><span class="mi">2</span><span class="p">].</span><span class="n">f2</span><span class="p">;</span>
+<span class="p">}</span>
+<span class="p">...</span>
+<span class="n">munger_struct</span> <span class="n">Array</span><span class="p">[</span><span class="mi">3</span><span class="p">];</span>
+<span class="p">...</span>
+<span class="n">munge</span><span class="p">(</span><span class="n">Array</span><span class="p">);</span>
+</pre></div>
+</div>
+<p>In this “C” example, the front end compiler (Clang) will generate three GEP
+instructions for the three indices through “P” in the assignment statement.  The
+function argument <tt class="docutils literal"><span class="pre">P</span></tt> will be the first operand of each of these GEP
+instructions.  The second operand indexes through that pointer.  The third
+operand will be the field offset into the <tt class="docutils literal"><span class="pre">struct</span> <span class="pre">munger_struct</span></tt> type, for
+either the <tt class="docutils literal"><span class="pre">f1</span></tt> or <tt class="docutils literal"><span class="pre">f2</span></tt> field. So, in LLVM assembly the <tt class="docutils literal"><span class="pre">munge</span></tt> function
+looks like:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="kt">void</span> <span class="nv">%munge</span><span class="p">(</span><span class="nv">%struct.munger_struct</span><span class="p">*</span> <span class="nv">%P</span><span class="p">)</span> <span class="p">{</span>
+<span class="nl">entry:</span>
+  <span class="nv">%tmp</span> <span class="p">=</span> <span class="k">getelementptr</span> <span class="nv">%struct.munger_struct</span><span class="p">*</span> <span class="nv">%P</span><span class="p">,</span> <span class="k">i32</span> <span class="m">1</span><span class="p">,</span> <span class="k">i32</span> <span class="m">0</span>
+  <span class="nv">%tmp</span> <span class="p">=</span> <span class="k">load</span> <span class="k">i32</span><span class="p">*</span> <span class="nv">%tmp</span>
+  <span class="nv">%tmp6</span> <span class="p">=</span> <span class="k">getelementptr</span> <span class="nv">%struct.munger_struct</span><span class="p">*</span> <span class="nv">%P</span><span class="p">,</span> <span class="k">i32</span> <span class="m">2</span><span class="p">,</span> <span class="k">i32</span> <span class="m">1</span>
+  <span class="nv">%tmp7</span> <span class="p">=</span> <span class="k">load</span> <span class="k">i32</span><span class="p">*</span> <span class="nv">%tmp6</span>
+  <span class="nv">%tmp8</span> <span class="p">=</span> <span class="k">add</span> <span class="k">i32</span> <span class="nv">%tmp7</span><span class="p">,</span> <span class="nv">%tmp</span>
+  <span class="nv">%tmp9</span> <span class="p">=</span> <span class="k">getelementptr</span> <span class="nv">%struct.munger_struct</span><span class="p">*</span> <span class="nv">%P</span><span class="p">,</span> <span class="k">i32</span> <span class="m">0</span><span class="p">,</span> <span class="k">i32</span> <span class="m">0</span>
+  <span class="k">store</span> <span class="k">i32</span> <span class="nv">%tmp8</span><span class="p">,</span> <span class="k">i32</span><span class="p">*</span> <span class="nv">%tmp9</span>
+  <span class="k">ret</span> <span class="kt">void</span>
+<span class="p">}</span>
+</pre></div>
+</div>
+<p>In each case the first operand is the pointer through which the GEP instruction
+starts. The same is true whether the first operand is an argument, allocated
+memory, or a global variable.</p>
+<p>To make this clear, let’s consider a more obtuse example:</p>
+<div class="highlight-llvm"><pre>%MyVar = uninitialized global i32
+...
+%idx1 = getelementptr i32* %MyVar, i64 0
+%idx2 = getelementptr i32* %MyVar, i64 1
+%idx3 = getelementptr i32* %MyVar, i64 2</pre>
+</div>
+<p>These GEP instructions are simply making address computations from the base
+address of <tt class="docutils literal"><span class="pre">MyVar</span></tt>.  They compute, as follows (using C syntax):</p>
+<div class="highlight-c++"><div class="highlight"><pre><span class="n">idx1</span> <span class="o">=</span> <span class="p">(</span><span class="kt">char</span><span class="o">*</span><span class="p">)</span> <span class="o">&</span><span class="n">MyVar</span> <span class="o">+</span> <span class="mi">0</span>
+<span class="n">idx2</span> <span class="o">=</span> <span class="p">(</span><span class="kt">char</span><span class="o">*</span><span class="p">)</span> <span class="o">&</span><span class="n">MyVar</span> <span class="o">+</span> <span class="mi">4</span>
+<span class="n">idx3</span> <span class="o">=</span> <span class="p">(</span><span class="kt">char</span><span class="o">*</span><span class="p">)</span> <span class="o">&</span><span class="n">MyVar</span> <span class="o">+</span> <span class="mi">8</span>
+</pre></div>
+</div>
+<p>Since the type <tt class="docutils literal"><span class="pre">i32</span></tt> is known to be four bytes long, the indices 0, 1 and 2
+translate into memory offsets of 0, 4, and 8, respectively. No memory is
+accessed to make these computations because the address of <tt class="docutils literal"><span class="pre">%MyVar</span></tt> is passed
+directly to the GEP instructions.</p>
+<p>The obtuse part of this example is in the cases of <tt class="docutils literal"><span class="pre">%idx2</span></tt> and <tt class="docutils literal"><span class="pre">%idx3</span></tt>. They
+result in the computation of addresses that point to memory past the end of the
+<tt class="docutils literal"><span class="pre">%MyVar</span></tt> global, which is only one <tt class="docutils literal"><span class="pre">i32</span></tt> long, not three <tt class="docutils literal"><span class="pre">i32</span></tt>s long.
+While this is legal in LLVM, it is inadvisable because any load or store with
+the pointer that results from these GEP instructions would produce undefined
+results.</p>
+</div>
+<div class="section" id="why-is-the-extra-0-index-required">
+<h3><a class="toc-backref" href="#id5">Why is the extra 0 index required?</a><a class="headerlink" href="#why-is-the-extra-0-index-required" title="Permalink to this headline">¶</a></h3>
+<p>Quick answer: there are no superfluous indices.</p>
+<p>This question arises most often when the GEP instruction is applied to a global
+variable which is always a pointer type. For example, consider this:</p>
+<div class="highlight-llvm"><pre>%MyStruct = uninitialized global { float*, i32 }
+...
+%idx = getelementptr { float*, i32 }* %MyStruct, i64 0, i32 1</pre>
+</div>
+<p>The GEP above yields an <tt class="docutils literal"><span class="pre">i32*</span></tt> by indexing the <tt class="docutils literal"><span class="pre">i32</span></tt> typed field of the
+structure <tt class="docutils literal"><span class="pre">%MyStruct</span></tt>. When people first look at it, they wonder why the <tt class="docutils literal"><span class="pre">i64</span>
+<span class="pre">0</span></tt> index is needed. However, a closer inspection of how globals and GEPs work
+reveals the need. Becoming aware of the following facts will dispel the
+confusion:</p>
+<ol class="arabic simple">
+<li>The type of <tt class="docutils literal"><span class="pre">%MyStruct</span></tt> is <em>not</em> <tt class="docutils literal"><span class="pre">{</span> <span class="pre">float*,</span> <span class="pre">i32</span> <span class="pre">}</span></tt> but rather <tt class="docutils literal"><span class="pre">{</span> <span class="pre">float*,</span>
+<span class="pre">i32</span> <span class="pre">}*</span></tt>. That is, <tt class="docutils literal"><span class="pre">%MyStruct</span></tt> is a pointer to a structure containing a
+pointer to a <tt class="docutils literal"><span class="pre">float</span></tt> and an <tt class="docutils literal"><span class="pre">i32</span></tt>.</li>
+<li>Point #1 is evidenced by noticing the type of the first operand of the GEP
+instruction (<tt class="docutils literal"><span class="pre">%MyStruct</span></tt>) which is <tt class="docutils literal"><span class="pre">{</span> <span class="pre">float*,</span> <span class="pre">i32</span> <span class="pre">}*</span></tt>.</li>
+<li>The first index, <tt class="docutils literal"><span class="pre">i64</span> <span class="pre">0</span></tt> is required to step over the global variable
+<tt class="docutils literal"><span class="pre">%MyStruct</span></tt>.  Since the first argument to the GEP instruction must always
+be a value of pointer type, the first index steps through that pointer. A
+value of 0 means 0 elements offset from that pointer.</li>
+<li>The second index, <tt class="docutils literal"><span class="pre">i32</span> <span class="pre">1</span></tt> selects the second field of the structure (the
+<tt class="docutils literal"><span class="pre">i32</span></tt>).</li>
+</ol>
+</div>
+<div class="section" id="what-is-dereferenced-by-gep">
+<h3><a class="toc-backref" href="#id6">What is dereferenced by GEP?</a><a class="headerlink" href="#what-is-dereferenced-by-gep" title="Permalink to this headline">¶</a></h3>
+<p>Quick answer: nothing.</p>
+<p>The GetElementPtr instruction dereferences nothing. That is, it doesn’t access
+memory in any way. That’s what the Load and Store instructions are for.  GEP is
+only involved in the computation of addresses. For example, consider this:</p>
+<div class="highlight-llvm"><pre>%MyVar = uninitialized global { [40 x i32 ]* }
+...
+%idx = getelementptr { [40 x i32]* }* %MyVar, i64 0, i32 0, i64 0, i64 17</pre>
+</div>
+<p>In this example, we have a global variable, <tt class="docutils literal"><span class="pre">%MyVar</span></tt> that is a pointer to a
+structure containing a pointer to an array of 40 ints. The GEP instruction seems
+to be accessing the 18th integer of the structure’s array of ints. However, this
+is actually an illegal GEP instruction. It won’t compile. The reason is that the
+pointer in the structure <em>must</em> be dereferenced in order to index into the
+array of 40 ints. Since the GEP instruction never accesses memory, it is
+illegal.</p>
+<p>In order to access the 18th integer in the array, you would need to do the
+following:</p>
+<div class="highlight-llvm"><pre>%idx = getelementptr { [40 x i32]* }* %, i64 0, i32 0
+%arr = load [40 x i32]** %idx
+%idx = getelementptr [40 x i32]* %arr, i64 0, i64 17</pre>
+</div>
+<p>In this case, we have to load the pointer in the structure with a load
+instruction before we can index into the array. If the example was changed to:</p>
+<div class="highlight-llvm"><pre>%MyVar = uninitialized global { [40 x i32 ] }
+...
+%idx = getelementptr { [40 x i32] }*, i64 0, i32 0, i64 17</pre>
+</div>
+<p>then everything works fine. In this case, the structure does not contain a
+pointer and the GEP instruction can index through the global variable, into the
+first field of the structure and access the 18th <tt class="docutils literal"><span class="pre">i32</span></tt> in the array there.</p>
+</div>
+<div class="section" id="why-don-t-gep-x-0-0-1-and-gep-x-1-alias">
+<h3><a class="toc-backref" href="#id7">Why don’t GEP x,0,0,1 and GEP x,1 alias?</a><a class="headerlink" href="#why-don-t-gep-x-0-0-1-and-gep-x-1-alias" title="Permalink to this headline">¶</a></h3>
+<p>Quick Answer: They compute different address locations.</p>
+<p>If you look at the first indices in these GEP instructions you find that they
+are different (0 and 1), therefore the address computation diverges with that
+index. Consider this example:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="nv">%MyVar</span> <span class="p">=</span> <span class="k">global</span> <span class="p">{</span> <span class="p">[</span><span class="m">10</span> <span class="k">x</span> <span class="k">i32</span> <span class="p">]</span> <span class="p">}</span>
+<span class="nv">%idx1</span> <span class="p">=</span> <span class="k">getelementptr</span> <span class="p">{</span> <span class="p">[</span><span class="m">10</span> <span class="k">x</span> <span class="k">i32</span> <span class="p">]</span> <span class="p">}*</span> <span class="nv">%MyVar</span><span class="p">,</span> <span class="k">i64</span> <span class="m">0</span><span class="p">,</span> <span class="k">i32</span> <span class="m">0</span><span class="p">,</span> <span class="k">i64</span> <span class="m">1</span>
+<span class="nv">%idx2</span> <span class="p">=</span> <span class="k">getelementptr</span> <span class="p">{</span> <span class="p">[</span><span class="m">10</span> <span class="k">x</span> <span class="k">i32</span> <span class="p">]</span> <span class="p">}*</span> <span class="nv">%MyVar</span><span class="p">,</span> <span class="k">i64</span> <span class="m">1</span>
+</pre></div>
+</div>
+<p>In this example, <tt class="docutils literal"><span class="pre">idx1</span></tt> computes the address of the second integer in the
+array that is in the structure in <tt class="docutils literal"><span class="pre">%MyVar</span></tt>, that is <tt class="docutils literal"><span class="pre">MyVar+4</span></tt>. The type of
+<tt class="docutils literal"><span class="pre">idx1</span></tt> is <tt class="docutils literal"><span class="pre">i32*</span></tt>. However, <tt class="docutils literal"><span class="pre">idx2</span></tt> computes the address of <em>the next</em>
+structure after <tt class="docutils literal"><span class="pre">%MyVar</span></tt>. The type of <tt class="docutils literal"><span class="pre">idx2</span></tt> is <tt class="docutils literal"><span class="pre">{</span> <span class="pre">[10</span> <span class="pre">x</span> <span class="pre">i32]</span> <span class="pre">}*</span></tt> and its
+value is equivalent to <tt class="docutils literal"><span class="pre">MyVar</span> <span class="pre">+</span> <span class="pre">40</span></tt> because it indexes past the ten 4-byte
+integers in <tt class="docutils literal"><span class="pre">MyVar</span></tt>. Obviously, in such a situation, the pointers don’t
+alias.</p>
+</div>
+<div class="section" id="why-do-gep-x-1-0-0-and-gep-x-1-alias">
+<h3><a class="toc-backref" href="#id8">Why do GEP x,1,0,0 and GEP x,1 alias?</a><a class="headerlink" href="#why-do-gep-x-1-0-0-and-gep-x-1-alias" title="Permalink to this headline">¶</a></h3>
+<p>Quick Answer: They compute the same address location.</p>
+<p>These two GEP instructions will compute the same address because indexing
+through the 0th element does not change the address. However, it does change the
+type. Consider this example:</p>
+<div class="highlight-llvm"><div class="highlight"><pre><span class="nv">%MyVar</span> <span class="p">=</span> <span class="k">global</span> <span class="p">{</span> <span class="p">[</span><span class="m">10</span> <span class="k">x</span> <span class="k">i32</span> <span class="p">]</span> <span class="p">}</span>
+<span class="nv">%idx1</span> <span class="p">=</span> <span class="k">getelementptr</span> <span class="p">{</span> <span class="p">[</span><span class="m">10</span> <span class="k">x</span> <span class="k">i32</span> <span class="p">]</span> <span class="p">}*</span> <span class="nv">%MyVar</span><span class="p">,</span> <span class="k">i64</span> <span class="m">1</span><span class="p">,</span> <span class="k">i32</span> <span class="m">0</span><span class="p">,</span> <span class="k">i64</span> <span class="m">0</span>
+<span class="nv">%idx2</span> <span class="p">=</span> <span class="k">getelementptr</span> <span class="p">{</span> <span class="p">[</span><span class="m">10</span> <span class="k">x</span> <span class="k">i32</span> <span class="p">]</span> <span class="p">}*</span> <span class="nv">%MyVar</span><span class="p">,</span> <span class="k">i64</span> <span class="m">1</span>
+</pre></div>
+</div>
+<p>In this example, the value of <tt class="docutils literal"><span class="pre">%idx1</span></tt> is <tt class="docutils literal"><span class="pre">%MyVar+40</span></tt> and its type is
+<tt class="docutils literal"><span class="pre">i32*</span></tt>. The value of <tt class="docutils literal"><span class="pre">%idx2</span></tt> is also <tt class="docutils literal"><span class="pre">MyVar+40</span></tt> but its type is <tt class="docutils literal"><span class="pre">{</span> <span class="pre">[10</span> <span class="pre">x</span>
+<span class="pre">i32]</span> <span class="pre">}*</span></tt>.</p>
+</div>
+<div class="section" id="can-gep-index-into-vector-elements">
+<h3><a class="toc-backref" href="#id9">Can GEP index into vector elements?</a><a class="headerlink" href="#can-gep-index-into-vector-elements" title="Permalink to this headline">¶</a></h3>
+<p>This hasn’t always been forcefully disallowed, though it’s not recommended.  It
+leads to awkward special cases in the optimizers, and fundamental inconsistency
+in the IR. In the future, it will probably be outright disallowed.</p>
+</div>
+<div class="section" id="what-effect-do-address-spaces-have-on-geps">
+<h3><a class="toc-backref" href="#id10">What effect do address spaces have on GEPs?</a><a class="headerlink" href="#what-effect-do-address-spaces-have-on-geps" title="Permalink to this headline">¶</a></h3>
+<p>None, except that the address space qualifier on the first operand pointer type
+always matches the address space qualifier on the result type.</p>
+</div>
+<div class="section" id="how-is-gep-different-from-ptrtoint-arithmetic-and-inttoptr">
+<h3><a class="toc-backref" href="#id11">How is GEP different from <tt class="docutils literal"><span class="pre">ptrtoint</span></tt>, arithmetic, and <tt class="docutils literal"><span class="pre">inttoptr</span></tt>?</a><a class="headerlink" href="#how-is-gep-different-from-ptrtoint-arithmetic-and-inttoptr" title="Permalink to this headline">¶</a></h3>
+<p>It’s very similar; there are only subtle differences.</p>
+<p>With ptrtoint, you have to pick an integer type. One approach is to pick i64;
+this is safe on everything LLVM supports (LLVM internally assumes pointers are
+never wider than 64 bits in many places), and the optimizer will actually narrow
+the i64 arithmetic down to the actual pointer size on targets which don’t
+support 64-bit arithmetic in most cases. However, there are some cases where it
+doesn’t do this. With GEP you can avoid this problem.</p>
+<p>Also, GEP carries additional pointer aliasing rules. It’s invalid to take a GEP
+from one object, address into a different separately allocated object, and
+dereference it. IR producers (front-ends) must follow this rule, and consumers
+(optimizers, specifically alias analysis) benefit from being able to rely on
+it. See the <a class="reference internal" href="#rules">Rules</a> section for more information.</p>
+<p>And, GEP is more concise in common cases.</p>
+<p>However, for the underlying integer computation implied, there is no
+difference.</p>
+</div>
+<div class="section" id="i-m-writing-a-backend-for-a-target-which-needs-custom-lowering-for-gep-how-do-i-do-this">
+<h3><a class="toc-backref" href="#id12">I’m writing a backend for a target which needs custom lowering for GEP. How do I do this?</a><a class="headerlink" href="#i-m-writing-a-backend-for-a-target-which-needs-custom-lowering-for-gep-how-do-i-do-this" title="Permalink to this headline">¶</a></h3>
+<p>You don’t. The integer computation implied by a GEP is target-independent.
+Typically what you’ll need to do is make your backend pattern-match expressions
+trees involving ADD, MUL, etc., which are what GEP is lowered into. This has the
+advantage of letting your code work correctly in more cases.</p>
+<p>GEP does use target-dependent parameters for the size and layout of data types,
+which targets can customize.</p>
+<p>If you require support for addressing units which are not 8 bits, you’ll need to
+fix a lot of code in the backend, with GEP lowering being only a small piece of
+the overall picture.</p>
+</div>
+<div class="section" id="how-does-vla-addressing-work-with-geps">
+<h3><a class="toc-backref" href="#id13">How does VLA addressing work with GEPs?</a><a class="headerlink" href="#how-does-vla-addressing-work-with-geps" title="Permalink to this headline">¶</a></h3>
+<p>GEPs don’t natively support VLAs. LLVM’s type system is entirely static, and GEP
+address computations are guided by an LLVM type.</p>
+<p>VLA indices can be implemented as linearized indices. For example, an expression
+like <tt class="docutils literal"><span class="pre">X[a][b][c]</span></tt>, must be effectively lowered into a form like
+<tt class="docutils literal"><span class="pre">X[a*m+b*n+c]</span></tt>, so that it appears to the GEP as a single-dimensional array
+reference.</p>
+<p>This means if you want to write an analysis which understands array indices and
+you want to support VLAs, your code will have to be prepared to reverse-engineer
+the linearization. One way to solve this problem is to use the ScalarEvolution
+library, which always presents VLA and non-VLA indexing in the same manner.</p>
+</div>
+</div>
+<div class="section" id="rules">
+<span id="id1"></span><h2><a class="toc-backref" href="#id14">Rules</a><a class="headerlink" href="#rules" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="what-happens-if-an-array-index-is-out-of-bounds">
+<h3><a class="toc-backref" href="#id15">What happens if an array index is out of bounds?</a><a class="headerlink" href="#what-happens-if-an-array-index-is-out-of-bounds" title="Permalink to this headline">¶</a></h3>
+<p>There are two senses in which an array index can be out of bounds.</p>
+<p>First, there’s the array type which comes from the (static) type of the first
+operand to the GEP. Indices greater than the number of elements in the
+corresponding static array type are valid. There is no problem with out of
+bounds indices in this sense. Indexing into an array only depends on the size of
+the array element, not the number of elements.</p>
+<p>A common example of how this is used is arrays where the size is not known.
+It’s common to use array types with zero length to represent these. The fact
+that the static type says there are zero elements is irrelevant; it’s perfectly
+valid to compute arbitrary element indices, as the computation only depends on
+the size of the array element, not the number of elements. Note that zero-sized
+arrays are not a special case here.</p>
+<p>This sense is unconnected with <tt class="docutils literal"><span class="pre">inbounds</span></tt> keyword. The <tt class="docutils literal"><span class="pre">inbounds</span></tt> keyword is
+designed to describe low-level pointer arithmetic overflow conditions, rather
+than high-level array indexing rules.</p>
+<p>Analysis passes which wish to understand array indexing should not assume that
+the static array type bounds are respected.</p>
+<p>The second sense of being out of bounds is computing an address that’s beyond
+the actual underlying allocated object.</p>
+<p>With the <tt class="docutils literal"><span class="pre">inbounds</span></tt> keyword, the result value of the GEP is undefined if the
+address is outside the actual underlying allocated object and not the address
+one-past-the-end.</p>
+<p>Without the <tt class="docutils literal"><span class="pre">inbounds</span></tt> keyword, there are no restrictions on computing
+out-of-bounds addresses. Obviously, performing a load or a store requires an
+address of allocated and sufficiently aligned memory. But the GEP itself is only
+concerned with computing addresses.</p>
+</div>
+<div class="section" id="can-array-indices-be-negative">
+<h3><a class="toc-backref" href="#id16">Can array indices be negative?</a><a class="headerlink" href="#can-array-indices-be-negative" title="Permalink to this headline">¶</a></h3>
+<p>Yes. This is basically a special case of array indices being out of bounds.</p>
+</div>
+<div class="section" id="can-i-compare-two-values-computed-with-geps">
+<h3><a class="toc-backref" href="#id17">Can I compare two values computed with GEPs?</a><a class="headerlink" href="#can-i-compare-two-values-computed-with-geps" title="Permalink to this headline">¶</a></h3>
+<p>Yes. If both addresses are within the same allocated object, or
+one-past-the-end, you’ll get the comparison result you expect. If either is
+outside of it, integer arithmetic wrapping may occur, so the comparison may not
+be meaningful.</p>
+</div>
+<div class="section" id="can-i-do-gep-with-a-different-pointer-type-than-the-type-of-the-underlying-object">
+<h3><a class="toc-backref" href="#id18">Can I do GEP with a different pointer type than the type of the underlying object?</a><a class="headerlink" href="#can-i-do-gep-with-a-different-pointer-type-than-the-type-of-the-underlying-object" title="Permalink to this headline">¶</a></h3>
+<p>Yes. There are no restrictions on bitcasting a pointer value to an arbitrary
+pointer type. The types in a GEP serve only to define the parameters for the
+underlying integer computation. They need not correspond with the actual type of
+the underlying object.</p>
+<p>Furthermore, loads and stores don’t have to use the same types as the type of
+the underlying object. Types in this context serve only to specify memory size
+and alignment. Beyond that there are merely a hint to the optimizer indicating
+how the value will likely be used.</p>
+</div>
+<div class="section" id="can-i-cast-an-object-s-address-to-integer-and-add-it-to-null">
+<h3><a class="toc-backref" href="#id19">Can I cast an object’s address to integer and add it to null?</a><a class="headerlink" href="#can-i-cast-an-object-s-address-to-integer-and-add-it-to-null" title="Permalink to this headline">¶</a></h3>
+<p>You can compute an address that way, but if you use GEP to do the add, you can’t
+use that pointer to actually access the object, unless the object is managed
+outside of LLVM.</p>
+<p>The underlying integer computation is sufficiently defined; null has a defined
+value — zero — and you can add whatever value you want to it.</p>
+<p>However, it’s invalid to access (load from or store to) an LLVM-aware object
+with such a pointer. This includes <tt class="docutils literal"><span class="pre">GlobalVariables</span></tt>, <tt class="docutils literal"><span class="pre">Allocas</span></tt>, and objects
+pointed to by noalias pointers.</p>
+<p>If you really need this functionality, you can do the arithmetic with explicit
+integer instructions, and use inttoptr to convert the result to an address. Most
+of GEP’s special aliasing rules do not apply to pointers computed from ptrtoint,
+arithmetic, and inttoptr sequences.</p>
+</div>
+<div class="section" id="can-i-compute-the-distance-between-two-objects-and-add-that-value-to-one-address-to-compute-the-other-address">
+<h3><a class="toc-backref" href="#id20">Can I compute the distance between two objects, and add that value to one address to compute the other address?</a><a class="headerlink" href="#can-i-compute-the-distance-between-two-objects-and-add-that-value-to-one-address-to-compute-the-other-address" title="Permalink to this headline">¶</a></h3>
+<p>As with arithmetic on null, you can use GEP to compute an address that way, but
+you can’t use that pointer to actually access the object if you do, unless the
+object is managed outside of LLVM.</p>
+<p>Also as above, ptrtoint and inttoptr provide an alternative way to do this which
+do not have this restriction.</p>
+</div>
+<div class="section" id="can-i-do-type-based-alias-analysis-on-llvm-ir">
+<h3><a class="toc-backref" href="#id21">Can I do type-based alias analysis on LLVM IR?</a><a class="headerlink" href="#can-i-do-type-based-alias-analysis-on-llvm-ir" title="Permalink to this headline">¶</a></h3>
+<p>You can’t do type-based alias analysis using LLVM’s built-in type system,
+because LLVM has no restrictions on mixing types in addressing, loads or stores.</p>
+<p>LLVM’s type-based alias analysis pass uses metadata to describe a different type
+system (such as the C type system), and performs type-based aliasing on top of
+that.  Further details are in the <a class="reference external" href="LangRef.html#tbaa">language reference</a>.</p>
+</div>
+<div class="section" id="what-happens-if-a-gep-computation-overflows">
+<h3><a class="toc-backref" href="#id22">What happens if a GEP computation overflows?</a><a class="headerlink" href="#what-happens-if-a-gep-computation-overflows" title="Permalink to this headline">¶</a></h3>
+<p>If the GEP lacks the <tt class="docutils literal"><span class="pre">inbounds</span></tt> keyword, the value is the result from
+evaluating the implied two’s complement integer computation. However, since
+there’s no guarantee of where an object will be allocated in the address space,
+such values have limited meaning.</p>
+<p>If the GEP has the <tt class="docutils literal"><span class="pre">inbounds</span></tt> keyword, the result value is undefined (a “trap
+value”) if the GEP overflows (i.e. wraps around the end of the address space).</p>
+<p>As such, there are some ramifications of this for inbounds GEPs: scales implied
+by array/vector/pointer indices are always known to be “nsw” since they are
+signed values that are scaled by the element size.  These values are also
+allowed to be negative (e.g. “<tt class="docutils literal"><span class="pre">gep</span> <span class="pre">i32</span> <span class="pre">*%P,</span> <span class="pre">i32</span> <span class="pre">-1</span></tt>”) but the pointer itself
+is logically treated as an unsigned value.  This means that GEPs have an
+asymmetric relation between the pointer base (which is treated as unsigned) and
+the offset applied to it (which is treated as signed). The result of the
+additions within the offset calculation cannot have signed overflow, but when
+applied to the base pointer, there can be signed overflow.</p>
+</div>
+<div class="section" id="how-can-i-tell-if-my-front-end-is-following-the-rules">
+<h3><a class="toc-backref" href="#id23">How can I tell if my front-end is following the rules?</a><a class="headerlink" href="#how-can-i-tell-if-my-front-end-is-following-the-rules" title="Permalink to this headline">¶</a></h3>
+<p>There is currently no checker for the getelementptr rules. Currently, the only
+way to do this is to manually check each place in your front-end where
+GetElementPtr operators are created.</p>
+<p>It’s not possible to write a checker which could find all rule violations
+statically. It would be possible to write a checker which works by instrumenting
+the code with dynamic checks though. Alternatively, it would be possible to
+write a static checker which catches a subset of possible problems. However, no
+such checker exists today.</p>
+</div>
+</div>
+<div class="section" id="rationale">
+<h2><a class="toc-backref" href="#id24">Rationale</a><a class="headerlink" href="#rationale" title="Permalink to this headline">¶</a></h2>
+<div class="section" id="why-is-gep-designed-this-way">
+<h3><a class="toc-backref" href="#id25">Why is GEP designed this way?</a><a class="headerlink" href="#why-is-gep-designed-this-way" title="Permalink to this headline">¶</a></h3>
+<p>The design of GEP has the following goals, in rough unofficial order of
+priority:</p>
+<ul class="simple">
+<li>Support C, C-like languages, and languages which can be conceptually lowered
+into C (this covers a lot).</li>
+<li>Support optimizations such as those that are common in C compilers. In
+particular, GEP is a cornerstone of LLVM’s <a class="reference external" href="LangRef.html#pointeraliasing">pointer aliasing
+model</a>.</li>
+<li>Provide a consistent method for computing addresses so that address
+computations don’t need to be a part of load and store instructions in the IR.</li>
+<li>Support non-C-like languages, to the extent that it doesn’t interfere with
+other goals.</li>
+<li>Minimize target-specific information in the IR.</li>
+</ul>
+</div>
+<div class="section" id="why-do-struct-member-indices-always-use-i32">
+<h3><a class="toc-backref" href="#id26">Why do struct member indices always use <tt class="docutils literal"><span class="pre">i32</span></tt>?</a><a class="headerlink" href="#why-do-struct-member-indices-always-use-i32" title="Permalink to this headline">¶</a></h3>
+<p>The specific type i32 is probably just a historical artifact, however it’s wide
+enough for all practical purposes, so there’s been no need to change it.  It
+doesn’t necessarily imply i32 address arithmetic; it’s just an identifier which
+identifies a field in a struct. Requiring that all struct indices be the same
+reduces the range of possibilities for cases where two GEPs are effectively the
+same but have distinct operand types.</p>
+</div>
+<div class="section" id="what-s-an-uglygep">
+<h3><a class="toc-backref" href="#id27">What’s an uglygep?</a><a class="headerlink" href="#what-s-an-uglygep" title="Permalink to this headline">¶</a></h3>
+<p>Some LLVM optimizers operate on GEPs by internally lowering them into more
+primitive integer expressions, which allows them to be combined with other
+integer expressions and/or split into multiple separate integer expressions. If
+they’ve made non-trivial changes, translating back into LLVM IR can involve
+reverse-engineering the structure of the addressing in order to fit it into the
+static type of the original first operand. It isn’t always possibly to fully
+reconstruct this structure; sometimes the underlying addressing doesn’t
+correspond with the static type at all. In such cases the optimizer instead will
+emit a GEP with the base pointer casted to a simple address-unit pointer, using
+the name “uglygep”. This isn’t pretty, but it’s just as valid, and it’s
+sufficient to preserve the pointer aliasing guarantees that GEP provides.</p>
+</div>
+</div>
+<div class="section" id="summary">
+<h2><a class="toc-backref" href="#id28">Summary</a><a class="headerlink" href="#summary" title="Permalink to this headline">¶</a></h2>
+<p>In summary, here’s some things to always remember about the GetElementPtr
+instruction:</p>
+<ol class="arabic simple">
+<li>The GEP instruction never accesses memory, it only provides pointer
+computations.</li>
+<li>The first operand to the GEP instruction is always a pointer and it must be
+indexed.</li>
+<li>There are no superfluous indices for the GEP instruction.</li>
+<li>Trailing zero indices are superfluous for pointer aliasing, but not for the
+types of the pointers.</li>
+<li>Leading zero indices are not superfluous for pointer aliasing nor the types
+of the pointers.</li>
+</ol>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="MCJITDesignAndImplementation.html" title="MCJIT Design and Implementation"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="YamlIO.html" title="YAML I/O"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/GettingStarted.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/GettingStarted.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/GettingStarted.html (added)
+++ www-releases/trunk/3.5.1/docs/GettingStarted.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,1347 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>Getting Started with the LLVM System — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="Getting Started with the LLVM System using Microsoft Visual Studio" href="GettingStartedVS.html" />
+    <link rel="prev" title="llvm-readobj - LLVM Object Reader" href="CommandGuide/llvm-readobj.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">
+      <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="GettingStartedVS.html" title="Getting Started with the LLVM System using Microsoft Visual Studio"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="CommandGuide/llvm-readobj.html" title="llvm-readobj - LLVM Object Reader"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="getting-started-with-the-llvm-system">
+<h1>Getting Started with the LLVM System<a class="headerlink" href="#getting-started-with-the-llvm-system" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#overview" id="id5">Overview</a></li>
+<li><a class="reference internal" href="#getting-started-quickly-a-summary" id="id6">Getting Started Quickly (A Summary)</a></li>
+<li><a class="reference internal" href="#requirements" id="id7">Requirements</a><ul>
+<li><a class="reference internal" href="#hardware" id="id8">Hardware</a></li>
+<li><a class="reference internal" href="#software" id="id9">Software</a></li>
+<li><a class="reference internal" href="#host-c-toolchain-both-compiler-and-standard-library" id="id10">Host C++ Toolchain, both Compiler and Standard Library</a><ul>
+<li><a class="reference internal" href="#getting-a-modern-host-c-toolchain" id="id11">Getting a Modern Host C++ Toolchain</a></li>
+</ul>
+</li>
+</ul>
+</li>
+<li><a class="reference internal" href="#getting-started-with-llvm" id="id12">Getting Started with LLVM</a><ul>
+<li><a class="reference internal" href="#terminology-and-notation" id="id13">Terminology and Notation</a></li>
+<li><a class="reference internal" href="#setting-up-your-environment" id="id14">Setting Up Your Environment</a></li>
+<li><a class="reference internal" href="#unpacking-the-llvm-archives" id="id15">Unpacking the LLVM Archives</a></li>
+<li><a class="reference internal" href="#checkout-llvm-from-subversion" id="id16">Checkout LLVM from Subversion</a></li>
+<li><a class="reference internal" href="#git-mirror" id="id17">Git Mirror</a><ul>
+<li><a class="reference internal" href="#sending-patches-with-git" id="id18">Sending patches with Git</a></li>
+<li><a class="reference internal" href="#for-developers-to-work-with-git-svn" id="id19">For developers to work with git-svn</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#local-llvm-configuration" id="id20">Local LLVM Configuration</a></li>
+<li><a class="reference internal" href="#compiling-the-llvm-suite-source-code" id="id21">Compiling the LLVM Suite Source Code</a></li>
+<li><a class="reference internal" href="#cross-compiling-llvm" id="id22">Cross-Compiling LLVM</a></li>
+<li><a class="reference internal" href="#the-location-of-llvm-object-files" id="id23">The Location of LLVM Object Files</a></li>
+<li><a class="reference internal" href="#optional-configuration-items" id="id24">Optional Configuration Items</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#general-layout" id="id25">Program Layout</a><ul>
+<li><a class="reference internal" href="#llvm-examples" id="id26"><tt class="docutils literal"><span class="pre">llvm/examples</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-include" id="id27"><tt class="docutils literal"><span class="pre">llvm/include</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-lib" id="id28"><tt class="docutils literal"><span class="pre">llvm/lib</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-projects" id="id29"><tt class="docutils literal"><span class="pre">llvm/projects</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-runtime" id="id30"><tt class="docutils literal"><span class="pre">llvm/runtime</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-test" id="id31"><tt class="docutils literal"><span class="pre">llvm/test</span></tt></a></li>
+<li><a class="reference internal" href="#test-suite" id="id32"><tt class="docutils literal"><span class="pre">test-suite</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-tools" id="id33"><tt class="docutils literal"><span class="pre">llvm/tools</span></tt></a></li>
+<li><a class="reference internal" href="#llvm-utils" id="id34"><tt class="docutils literal"><span class="pre">llvm/utils</span></tt></a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#an-example-using-the-llvm-tool-chain" id="id35">An Example Using the LLVM Tool Chain</a><ul>
+<li><a class="reference internal" href="#example-with-clang" id="id36">Example with clang</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#common-problems" id="id37">Common Problems</a></li>
+<li><a class="reference internal" href="#links" id="id38">Links</a></li>
+</ul>
+</div>
+<div class="section" id="overview">
+<h2><a class="toc-backref" href="#id5">Overview</a><a class="headerlink" href="#overview" title="Permalink to this headline">¶</a></h2>
+<p>Welcome to LLVM! In order to get started, you first need to know some basic
+information.</p>
+<p>First, LLVM comes in three pieces. The first piece is the LLVM suite. This
+contains all of the tools, libraries, and header files needed to use LLVM.  It
+contains an assembler, disassembler, bitcode analyzer and bitcode optimizer.  It
+also contains basic regression tests that can be used to test the LLVM tools and
+the Clang front end.</p>
+<p>The second piece is the <a class="reference external" href="http://clang.llvm.org/">Clang</a> front end.  This
+component compiles C, C++, Objective C, and Objective C++ code into LLVM
+bitcode. Once compiled into LLVM bitcode, a program can be manipulated with the
+LLVM tools from the LLVM suite.</p>
+<p>There is a third, optional piece called Test Suite.  It is a suite of programs
+with a testing harness that can be used to further test LLVM’s functionality
+and performance.</p>
+</div>
+<div class="section" id="getting-started-quickly-a-summary">
+<h2><a class="toc-backref" href="#id6">Getting Started Quickly (A Summary)</a><a class="headerlink" href="#getting-started-quickly-a-summary" title="Permalink to this headline">¶</a></h2>
+<p>The LLVM Getting Started documentation may be out of date.  So, the <a class="reference external" href="http://clang.llvm.org/get_started.html">Clang
+Getting Started</a> page might also be a
+good place to start.</p>
+<p>Here’s the short story for getting up and running quickly with LLVM:</p>
+<ol class="arabic simple">
+<li>Read the documentation.</li>
+<li>Read the documentation.</li>
+<li>Remember that you were warned twice about reading the documentation.</li>
+<li>Checkout LLVM:<ul>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">where-you-want-llvm-to-live</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">svn</span> <span class="pre">co</span> <span class="pre">http://llvm.org/svn/llvm-project/llvm/trunk</span> <span class="pre">llvm</span></tt></li>
+</ul>
+</li>
+<li>Checkout Clang:<ul>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">where-you-want-llvm-to-live</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">llvm/tools</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">svn</span> <span class="pre">co</span> <span class="pre">http://llvm.org/svn/llvm-project/cfe/trunk</span> <span class="pre">clang</span></tt></li>
+</ul>
+</li>
+<li>Checkout Compiler-RT:<ul>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">where-you-want-llvm-to-live</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">llvm/projects</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">svn</span> <span class="pre">co</span> <span class="pre">http://llvm.org/svn/llvm-project/compiler-rt/trunk</span> <span class="pre">compiler-rt</span></tt></li>
+</ul>
+</li>
+<li>Get the Test Suite Source Code <strong>[Optional]</strong><ul>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">where-you-want-llvm-to-live</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">llvm/projects</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">svn</span> <span class="pre">co</span> <span class="pre">http://llvm.org/svn/llvm-project/test-suite/trunk</span> <span class="pre">test-suite</span></tt></li>
+</ul>
+</li>
+<li>Configure and build LLVM and Clang:<ul>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">where-you-want-to-build-llvm</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">mkdir</span> <span class="pre">build</span></tt> (for building without polluting the source dir)</li>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">build</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">../llvm/configure</span> <span class="pre">[options]</span></tt>
+Some common options:<ul>
+<li><tt class="docutils literal"><span class="pre">--prefix=directory</span></tt> — Specify for <em>directory</em> the full pathname of
+where you want the LLVM tools and libraries to be installed (default
+<tt class="docutils literal"><span class="pre">/usr/local</span></tt>).</li>
+<li><tt class="docutils literal"><span class="pre">--enable-optimized</span></tt> — Compile with optimizations enabled (default
+is NO).</li>
+<li><tt class="docutils literal"><span class="pre">--enable-assertions</span></tt> — Compile with assertion checks enabled
+(default is YES).</li>
+</ul>
+</li>
+<li><tt class="docutils literal"><span class="pre">make</span> <span class="pre">[-j]</span></tt> — The <tt class="docutils literal"><span class="pre">-j</span></tt> specifies the number of jobs (commands) to run
+simultaneously.  This builds both LLVM and Clang for Debug+Asserts mode.
+The <tt class="docutils literal"><span class="pre">--enable-optimized</span></tt> configure option is used to specify a Release
+build.</li>
+<li><tt class="docutils literal"><span class="pre">make</span> <span class="pre">check-all</span></tt> — This run the regression tests to ensure everything
+is in working order.</li>
+<li>It is also possible to use <a class="reference external" href="CMake.html">CMake</a> instead of the makefiles.
+With CMake it is possible to generate project files for several IDEs:
+Xcode, Eclipse CDT4, CodeBlocks, Qt-Creator (use the CodeBlocks
+generator), KDevelop3.</li>
+<li>If you get an “internal compiler error (ICE)” or test failures, see
+<cite>below</cite>.</li>
+</ul>
+</li>
+</ol>
+<p>Consult the <a class="reference internal" href="#getting-started-with-llvm">Getting Started with LLVM</a> section for detailed information on
+configuring and compiling LLVM.  See <a class="reference internal" href="#setting-up-your-environment">Setting Up Your Environment</a> for tips
+that simplify working with the Clang front end and LLVM tools.  Go to <a class="reference internal" href="#program-layout">Program
+Layout</a> to learn about the layout of the source code tree.</p>
+</div>
+<div class="section" id="requirements">
+<h2><a class="toc-backref" href="#id7">Requirements</a><a class="headerlink" href="#requirements" title="Permalink to this headline">¶</a></h2>
+<p>Before you begin to use the LLVM system, review the requirements given below.
+This may save you some trouble by knowing ahead of time what hardware and
+software you will need.</p>
+<div class="section" id="hardware">
+<h3><a class="toc-backref" href="#id8">Hardware</a><a class="headerlink" href="#hardware" title="Permalink to this headline">¶</a></h3>
+<p>LLVM is known to work on the following host platforms:</p>
+<table border="1" class="docutils">
+<colgroup>
+<col width="35%" />
+<col width="40%" />
+<col width="25%" />
+</colgroup>
+<thead valign="bottom">
+<tr class="row-odd"><th class="head">OS</th>
+<th class="head">Arch</th>
+<th class="head">Compilers</th>
+</tr>
+</thead>
+<tbody valign="top">
+<tr class="row-even"><td>AuroraUX</td>
+<td>x86<sup>1</sup></td>
+<td>GCC</td>
+</tr>
+<tr class="row-odd"><td>Linux</td>
+<td>x86<sup>1</sup></td>
+<td>GCC, Clang</td>
+</tr>
+<tr class="row-even"><td>Linux</td>
+<td>amd64</td>
+<td>GCC, Clang</td>
+</tr>
+<tr class="row-odd"><td>Linux</td>
+<td>ARM<sup>4</sup></td>
+<td>GCC, Clang</td>
+</tr>
+<tr class="row-even"><td>Linux</td>
+<td>PowerPC</td>
+<td>GCC, Clang</td>
+</tr>
+<tr class="row-odd"><td>Solaris</td>
+<td>V9 (Ultrasparc)</td>
+<td>GCC</td>
+</tr>
+<tr class="row-even"><td>FreeBSD</td>
+<td>x86<sup>1</sup></td>
+<td>GCC, Clang</td>
+</tr>
+<tr class="row-odd"><td>FreeBSD</td>
+<td>amd64</td>
+<td>GCC, Clang</td>
+</tr>
+<tr class="row-even"><td>MacOS X<sup>2</sup></td>
+<td>PowerPC</td>
+<td>GCC</td>
+</tr>
+<tr class="row-odd"><td>MacOS X</td>
+<td>x86</td>
+<td>GCC, Clang</td>
+</tr>
+<tr class="row-even"><td>Cygwin/Win32</td>
+<td>x86<sup>1, 3</sup></td>
+<td>GCC</td>
+</tr>
+<tr class="row-odd"><td>Windows</td>
+<td>x86<sup>1</sup></td>
+<td>Visual Studio</td>
+</tr>
+<tr class="row-even"><td>Windows x64</td>
+<td>x86-64</td>
+<td>Visual Studio</td>
+</tr>
+</tbody>
+</table>
+<div class="admonition note">
+<p class="first admonition-title">Note</p>
+<ol class="last arabic simple">
+<li>Code generation supported for Pentium processors and up</li>
+<li>Code generation supported for 32-bit ABI only</li>
+<li>To use LLVM modules on Win32-based system, you may configure LLVM
+with <tt class="docutils literal"><span class="pre">--enable-shared</span></tt>.</li>
+<li>MCJIT not working well pre-v7, old JIT engine not supported any more.</li>
+</ol>
+</div>
+<p>Note that you will need about 1-3 GB of space for a full LLVM build in Debug
+mode, depending on the system (it is so large because of all the debugging
+information and the fact that the libraries are statically linked into multiple
+tools).  If you do not need many of the tools and you are space-conscious, you
+can pass <tt class="docutils literal"><span class="pre">ONLY_TOOLS="tools</span> <span class="pre">you</span> <span class="pre">need"</span></tt> to make.  The Release build requires
+considerably less space.</p>
+<p>The LLVM suite <em>may</em> compile on other platforms, but it is not guaranteed to do
+so.  If compilation is successful, the LLVM utilities should be able to
+assemble, disassemble, analyze, and optimize LLVM bitcode.  Code generation
+should work as well, although the generated native code may not work on your
+platform.</p>
+</div>
+<div class="section" id="software">
+<h3><a class="toc-backref" href="#id9">Software</a><a class="headerlink" href="#software" title="Permalink to this headline">¶</a></h3>
+<p>Compiling LLVM requires that you have several software packages installed. The
+table below lists those required packages. The Package column is the usual name
+for the software package that LLVM depends on. The Version column provides
+“known to work” versions of the package. The Notes column describes how LLVM
+uses the package and provides other details.</p>
+<table border="1" class="docutils">
+<colgroup>
+<col width="52%" />
+<col width="11%" />
+<col width="38%" />
+</colgroup>
+<thead valign="bottom">
+<tr class="row-odd"><th class="head">Package</th>
+<th class="head">Version</th>
+<th class="head">Notes</th>
+</tr>
+</thead>
+<tbody valign="top">
+<tr class="row-even"><td><a class="reference external" href="http://savannah.gnu.org/projects/make">GNU Make</a></td>
+<td>3.79, 3.79.1</td>
+<td>Makefile/build processor</td>
+</tr>
+<tr class="row-odd"><td><a class="reference external" href="http://gcc.gnu.org/">GCC</a></td>
+<td>>=4.7.0</td>
+<td>C/C++ compiler<sup>1</sup></td>
+</tr>
+<tr class="row-even"><td><a class="reference external" href="http://www.python.org/">python</a></td>
+<td>>=2.5</td>
+<td>Automated test suite<sup>2</sup></td>
+</tr>
+<tr class="row-odd"><td><a class="reference external" href="http://savannah.gnu.org/projects/m4">GNU M4</a></td>
+<td>1.4</td>
+<td>Macro processor for configuration<sup>3</sup></td>
+</tr>
+<tr class="row-even"><td><a class="reference external" href="http://www.gnu.org/software/autoconf/">GNU Autoconf</a></td>
+<td>2.60</td>
+<td>Configuration script builder<sup>3</sup></td>
+</tr>
+<tr class="row-odd"><td><a class="reference external" href="http://www.gnu.org/software/automake/">GNU Automake</a></td>
+<td>1.9.6</td>
+<td>aclocal macro generator<sup>3</sup></td>
+</tr>
+<tr class="row-even"><td><a class="reference external" href="http://savannah.gnu.org/projects/libtool">libtool</a></td>
+<td>1.5.22</td>
+<td>Shared library manager<sup>3</sup></td>
+</tr>
+<tr class="row-odd"><td><a class="reference external" href="http://zlib.net">zlib</a></td>
+<td>>=1.2.3.4</td>
+<td>Compression library<sup>4</sup></td>
+</tr>
+</tbody>
+</table>
+<div class="admonition note">
+<p class="first admonition-title">Note</p>
+<ol class="last arabic simple">
+<li>Only the C and C++ languages are needed so there’s no need to build the
+other languages for LLVM’s purposes. See <cite>below</cite> for specific version
+info.</li>
+<li>Only needed if you want to run the automated test suite in the
+<tt class="docutils literal"><span class="pre">llvm/test</span></tt> directory.</li>
+<li>If you want to make changes to the configure scripts, you will need GNU
+autoconf (2.60), and consequently, GNU M4 (version 1.4 or higher). You
+will also need automake (1.9.6). We only use aclocal from that package.</li>
+<li>Optional, adds compression / uncompression capabilities to selected LLVM
+tools.</li>
+</ol>
+</div>
+<p>Additionally, your compilation host is expected to have the usual plethora of
+Unix utilities. Specifically:</p>
+<ul class="simple">
+<li><strong>ar</strong> — archive library builder</li>
+<li><strong>bzip2</strong> — bzip2 command for distribution generation</li>
+<li><strong>bunzip2</strong> — bunzip2 command for distribution checking</li>
+<li><strong>chmod</strong> — change permissions on a file</li>
+<li><strong>cat</strong> — output concatenation utility</li>
+<li><strong>cp</strong> — copy files</li>
+<li><strong>date</strong> — print the current date/time</li>
+<li><strong>echo</strong> — print to standard output</li>
+<li><strong>egrep</strong> — extended regular expression search utility</li>
+<li><strong>find</strong> — find files/dirs in a file system</li>
+<li><strong>grep</strong> — regular expression search utility</li>
+<li><strong>gzip</strong> — gzip command for distribution generation</li>
+<li><strong>gunzip</strong> — gunzip command for distribution checking</li>
+<li><strong>install</strong> — install directories/files</li>
+<li><strong>mkdir</strong> — create a directory</li>
+<li><strong>mv</strong> — move (rename) files</li>
+<li><strong>ranlib</strong> — symbol table builder for archive libraries</li>
+<li><strong>rm</strong> — remove (delete) files and directories</li>
+<li><strong>sed</strong> — stream editor for transforming output</li>
+<li><strong>sh</strong> — Bourne shell for make build scripts</li>
+<li><strong>tar</strong> — tape archive for distribution generation</li>
+<li><strong>test</strong> — test things in file system</li>
+<li><strong>unzip</strong> — unzip command for distribution checking</li>
+<li><strong>zip</strong> — zip command for distribution generation</li>
+</ul>
+</div>
+<div class="section" id="host-c-toolchain-both-compiler-and-standard-library">
+<span id="check-here"></span><span id="below"></span><h3><a class="toc-backref" href="#id10">Host C++ Toolchain, both Compiler and Standard Library</a><a class="headerlink" href="#host-c-toolchain-both-compiler-and-standard-library" title="Permalink to this headline">¶</a></h3>
+<p>LLVM is very demanding of the host C++ compiler, and as such tends to expose
+bugs in the compiler. We are also planning to follow improvements and
+developments in the C++ language and library reasonably closely. As such, we
+require a modern host C++ toolchain, both compiler and standard library, in
+order to build LLVM.</p>
+<p>For the most popular host toolchains we check for specific minimum versions in
+our build systems:</p>
+<ul class="simple">
+<li>Clang 3.1</li>
+<li>GCC 4.7</li>
+<li>Visual Studio 2012</li>
+</ul>
+<p>Anything older than these toolchains <em>may</em> work, but will require forcing the
+build system with a special option and is not really a supported host platform.
+Also note that older versions of these compilers have often crashed or
+miscompiled LLVM.</p>
+<p>For less widely used host toolchains such as ICC or xlC, be aware that a very
+recent version may be required to support all of the C++ features used in LLVM.</p>
+<p>We track certain versions of software that are <em>known</em> to fail when used as
+part of the host toolchain. These even include linkers at times.</p>
+<p><strong>GCC 4.6.3 on ARM</strong>: Miscompiles <tt class="docutils literal"><span class="pre">llvm-readobj</span></tt> at <tt class="docutils literal"><span class="pre">-O3</span></tt>. A test failure
+in <tt class="docutils literal"><span class="pre">test/Object/readobj-shared-object.test</span></tt> is one symptom of the problem.</p>
+<p><strong>GNU ld 2.16.X</strong>. Some 2.16.X versions of the ld linker will produce very long
+warning messages complaining that some “<tt class="docutils literal"><span class="pre">.gnu.linkonce.t.*</span></tt>” symbol was
+defined in a discarded section. You can safely ignore these messages as they are
+erroneous and the linkage is correct.  These messages disappear using ld 2.17.</p>
+<p><strong>GNU binutils 2.17</strong>: Binutils 2.17 contains <a class="reference external" href="http://sourceware.org/bugzilla/show_bug.cgi?id=3111">a bug</a> which causes huge link
+times (minutes instead of seconds) when building LLVM.  We recommend upgrading
+to a newer version (2.17.50.0.4 or later).</p>
+<p><strong>GNU Binutils 2.19.1 Gold</strong>: This version of Gold contained <a class="reference external" href="http://sourceware.org/bugzilla/show_bug.cgi?id=9836">a bug</a> which causes
+intermittent failures when building LLVM with position independent code.  The
+symptom is an error about cyclic dependencies.  We recommend upgrading to a
+newer version of Gold.</p>
+<p><strong>Clang 3.0 with libstdc++ 4.7.x</strong>: a few Linux distributions (Ubuntu 12.10,
+Fedora 17) have both Clang 3.0 and libstdc++ 4.7 in their repositories.  Clang
+3.0 does not implement a few builtins that are used in this library.  We
+recommend using the system GCC to compile LLVM and Clang in this case.</p>
+<p><strong>Clang 3.0 on Mageia 2</strong>.  There’s a packaging issue: Clang can not find at
+least some (<tt class="docutils literal"><span class="pre">cxxabi.h</span></tt>) libstdc++ headers.</p>
+<p><strong>Clang in C++11 mode and libstdc++ 4.7.2</strong>.  This version of libstdc++
+contained <a class="reference external" href="http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53841">a bug</a> which
+causes Clang to refuse to compile condition_variable header file.  At the time
+of writing, this breaks LLD build.</p>
+<div class="section" id="getting-a-modern-host-c-toolchain">
+<h4><a class="toc-backref" href="#id11">Getting a Modern Host C++ Toolchain</a><a class="headerlink" href="#getting-a-modern-host-c-toolchain" title="Permalink to this headline">¶</a></h4>
+<p>This section mostly applies to Linux and older BSDs. On Mac OS X, you should
+have a sufficiently modern Xcode, or you will likely need to upgrade until you
+do. On Windows, just use Visual Studio 2012 as the host compiler, it is
+explicitly supported and widely available. FreeBSD 10.0 and newer have a modern
+Clang as the system compiler.</p>
+<p>However, some Linux distributions and some other or older BSDs sometimes have
+extremely old versions of GCC. These steps attempt to help you upgrade you
+compiler even on such a system. However, if at all possible, we encourage you
+to use a recent version of a distribution with a modern system compiler that
+meets these requirements. Note that it is tempting to to install a prior
+version of Clang and libc++ to be the host compiler, however libc++ was not
+well tested or set up to build on Linux until relatively recently. As
+a consequence, this guide suggests just using libstdc++ and a modern GCC as the
+initial host in a bootstrap, and then using Clang (and potentially libc++).</p>
+<p>The first step is to get a recent GCC toolchain installed. The most common
+distribution on which users have struggled with the version requirements is
+Ubuntu Precise, 12.04 LTS. For this distribution, one easy option is to install
+the <a class="reference external" href="https://launchpad.net/~ubuntu-toolchain-r/+archive/test">toolchain testing PPA</a> and use it to install a modern GCC. There is
+a really nice discussions of this on the <a class="reference external" href="http://askubuntu.com/questions/271388/how-to-install-gcc-4-8-in-ubuntu-12-04-from-the-terminal">ask ubuntu stack exchange</a>. However,
+not all users can use PPAs and there are many other distributions, so it may be
+necessary (or just useful, if you’re here you <em>are</em> doing compiler development
+after all) to build and install GCC from source. It is also quite easy to do
+these days.</p>
+<p>Easy steps for installing GCC 4.8.2:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> wget ftp://ftp.gnu.org/gnu/gcc/gcc-4.8.2/gcc-4.8.2.tar.bz2
+<span class="gp">%</span> tar -xvjf gcc-4.8.2.tar.bz2
+<span class="gp">%</span> <span class="nb">cd </span>gcc-4.8.2
+<span class="gp">%</span> ./contrib/download_prerequisites
+<span class="gp">%</span> <span class="nb">cd</span> ..
+<span class="gp">%</span> mkdir gcc-4.8.2-build
+<span class="gp">%</span> <span class="nb">cd </span>gcc-4.8.2-build
+<span class="gp">%</span> <span class="nv">$PWD</span>/../gcc-4.8.2/configure --prefix<span class="o">=</span><span class="nv">$HOME</span>/toolchains --enable-languages<span class="o">=</span>c,c++
+<span class="gp">%</span> make -j<span class="k">$(</span>nproc<span class="k">)</span>
+<span class="gp">%</span> make install
+</pre></div>
+</div>
+<p>For more details, check out the excellent <a class="reference external" href="http://gcc.gnu.org/wiki/InstallingGCC">GCC wiki entry</a>, where I got most
+of this information from.</p>
+<p>Once you have a GCC toolchain, use it as your host compiler. Things should
+generally “just work”. You may need to pass a special linker flag,
+<tt class="docutils literal"><span class="pre">-Wl,-rpath,$HOME/toolchains/lib</span></tt> or some variant thereof to get things to
+find the libstdc++ DSO in this toolchain.</p>
+<p>When you build Clang, you will need to give <em>it</em> access to modern C++11
+standard library in order to use it as your new host in part of a bootstrap.
+There are two easy ways to do this, either build (and install) libc++ along
+with Clang and then use it with the <tt class="docutils literal"><span class="pre">-stdlib=libc++</span></tt> compile and link flag,
+or install Clang into the same prefix (<tt class="docutils literal"><span class="pre">$HOME/toolchains</span></tt> above) as GCC.
+Clang will look within its own prefix for libstdc++ and use it if found. You
+can also add an explicit prefix for Clang to look in for a GCC toolchain with
+the <tt class="docutils literal"><span class="pre">--gcc-toolchain=/opt/my/gcc/prefix</span></tt> flag, passing it to both compile and
+link commands when using your just-built-Clang to bootstrap.</p>
+</div>
+</div>
+</div>
+<div class="section" id="getting-started-with-llvm">
+<span id="id1"></span><h2><a class="toc-backref" href="#id12">Getting Started with LLVM</a><a class="headerlink" href="#getting-started-with-llvm" title="Permalink to this headline">¶</a></h2>
+<p>The remainder of this guide is meant to get you up and running with LLVM and to
+give you some basic information about the LLVM environment.</p>
+<p>The later sections of this guide describe the <a class="reference internal" href="#general-layout">general layout</a> of the LLVM
+source tree, a <a class="reference internal" href="#simple-example">simple example</a> using the LLVM tool chain, and <a class="reference internal" href="#links">links</a> to find
+more information about LLVM or to get help via e-mail.</p>
+<div class="section" id="terminology-and-notation">
+<h3><a class="toc-backref" href="#id13">Terminology and Notation</a><a class="headerlink" href="#terminology-and-notation" title="Permalink to this headline">¶</a></h3>
+<p>Throughout this manual, the following names are used to denote paths specific to
+the local system and working environment.  <em>These are not environment variables
+you need to set but just strings used in the rest of this document below</em>.  In
+any of the examples below, simply replace each of these names with the
+appropriate pathname on your local system.  All these paths are absolute:</p>
+<p><tt class="docutils literal"><span class="pre">SRC_ROOT</span></tt></p>
+<blockquote>
+<div>This is the top level directory of the LLVM source tree.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">OBJ_ROOT</span></tt></p>
+<blockquote>
+<div>This is the top level directory of the LLVM object tree (i.e. the tree where
+object files and compiled programs will be placed.  It can be the same as
+SRC_ROOT).</div></blockquote>
+</div>
+<div class="section" id="setting-up-your-environment">
+<span id="id2"></span><h3><a class="toc-backref" href="#id14">Setting Up Your Environment</a><a class="headerlink" href="#setting-up-your-environment" title="Permalink to this headline">¶</a></h3>
+<p>In order to compile and use LLVM, you may need to set some environment
+variables.</p>
+<p><tt class="docutils literal"><span class="pre">LLVM_LIB_SEARCH_PATH=/path/to/your/bitcode/libs</span></tt></p>
+<blockquote>
+<div>[Optional] This environment variable helps LLVM linking tools find the
+locations of your bitcode libraries. It is provided only as a convenience
+since you can specify the paths using the -L options of the tools and the
+C/C++ front-end will automatically use the bitcode files installed in its
+<tt class="docutils literal"><span class="pre">lib</span></tt> directory.</div></blockquote>
+</div>
+<div class="section" id="unpacking-the-llvm-archives">
+<h3><a class="toc-backref" href="#id15">Unpacking the LLVM Archives</a><a class="headerlink" href="#unpacking-the-llvm-archives" title="Permalink to this headline">¶</a></h3>
+<p>If you have the LLVM distribution, you will need to unpack it before you can
+begin to compile it.  LLVM is distributed as a set of two files: the LLVM suite
+and the LLVM GCC front end compiled for your platform.  There is an additional
+test suite that is optional.  Each file is a TAR archive that is compressed with
+the gzip program.</p>
+<p>The files are as follows, with <em>x.y</em> marking the version number:</p>
+<p><tt class="docutils literal"><span class="pre">llvm-x.y.tar.gz</span></tt></p>
+<blockquote>
+<div>Source release for the LLVM libraries and tools.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm-test-x.y.tar.gz</span></tt></p>
+<blockquote>
+<div>Source release for the LLVM test-suite.</div></blockquote>
+</div>
+<div class="section" id="checkout-llvm-from-subversion">
+<span id="checkout"></span><h3><a class="toc-backref" href="#id16">Checkout LLVM from Subversion</a><a class="headerlink" href="#checkout-llvm-from-subversion" title="Permalink to this headline">¶</a></h3>
+<p>If you have access to our Subversion repository, you can get a fresh copy of the
+entire source code.  All you need to do is check it out from Subversion as
+follows:</p>
+<ul class="simple">
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">where-you-want-llvm-to-live</span></tt></li>
+<li>Read-Only: <tt class="docutils literal"><span class="pre">svn</span> <span class="pre">co</span> <span class="pre">http://llvm.org/svn/llvm-project/llvm/trunk</span> <span class="pre">llvm</span></tt></li>
+<li>Read-Write:<tt class="docutils literal"><span class="pre">svn</span> <span class="pre">co</span> <span class="pre">https://user@llvm.org/svn/llvm-project/llvm/trunk</span> <span class="pre">llvm</span></tt></li>
+</ul>
+<p>This will create an ‘<tt class="docutils literal"><span class="pre">llvm</span></tt>‘ directory in the current directory and fully
+populate it with the LLVM source code, Makefiles, test directories, and local
+copies of documentation files.</p>
+<p>If you want to get a specific release (as opposed to the most recent revision),
+you can checkout it from the ‘<tt class="docutils literal"><span class="pre">tags</span></tt>‘ directory (instead of ‘<tt class="docutils literal"><span class="pre">trunk</span></tt>‘). The
+following releases are located in the following subdirectories of the ‘<tt class="docutils literal"><span class="pre">tags</span></tt>‘
+directory:</p>
+<ul class="simple">
+<li>Release 3.4: <strong>RELEASE_34/final</strong></li>
+<li>Release 3.3: <strong>RELEASE_33/final</strong></li>
+<li>Release 3.2: <strong>RELEASE_32/final</strong></li>
+<li>Release 3.1: <strong>RELEASE_31/final</strong></li>
+<li>Release 3.0: <strong>RELEASE_30/final</strong></li>
+<li>Release 2.9: <strong>RELEASE_29/final</strong></li>
+<li>Release 2.8: <strong>RELEASE_28</strong></li>
+<li>Release 2.7: <strong>RELEASE_27</strong></li>
+<li>Release 2.6: <strong>RELEASE_26</strong></li>
+<li>Release 2.5: <strong>RELEASE_25</strong></li>
+<li>Release 2.4: <strong>RELEASE_24</strong></li>
+<li>Release 2.3: <strong>RELEASE_23</strong></li>
+<li>Release 2.2: <strong>RELEASE_22</strong></li>
+<li>Release 2.1: <strong>RELEASE_21</strong></li>
+<li>Release 2.0: <strong>RELEASE_20</strong></li>
+<li>Release 1.9: <strong>RELEASE_19</strong></li>
+<li>Release 1.8: <strong>RELEASE_18</strong></li>
+<li>Release 1.7: <strong>RELEASE_17</strong></li>
+<li>Release 1.6: <strong>RELEASE_16</strong></li>
+<li>Release 1.5: <strong>RELEASE_15</strong></li>
+<li>Release 1.4: <strong>RELEASE_14</strong></li>
+<li>Release 1.3: <strong>RELEASE_13</strong></li>
+<li>Release 1.2: <strong>RELEASE_12</strong></li>
+<li>Release 1.1: <strong>RELEASE_11</strong></li>
+<li>Release 1.0: <strong>RELEASE_1</strong></li>
+</ul>
+<p>If you would like to get the LLVM test suite (a separate package as of 1.4), you
+get it from the Subversion repository:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nb">cd </span>llvm/projects
+<span class="gp">%</span> svn co http://llvm.org/svn/llvm-project/test-suite/trunk <span class="nb">test</span>-suite
+</pre></div>
+</div>
+<p>By placing it in the <tt class="docutils literal"><span class="pre">llvm/projects</span></tt>, it will be automatically configured by
+the LLVM configure script as well as automatically updated when you run <tt class="docutils literal"><span class="pre">svn</span>
+<span class="pre">update</span></tt>.</p>
+</div>
+<div class="section" id="git-mirror">
+<h3><a class="toc-backref" href="#id17">Git Mirror</a><a class="headerlink" href="#git-mirror" title="Permalink to this headline">¶</a></h3>
+<p>Git mirrors are available for a number of LLVM subprojects. These mirrors sync
+automatically with each Subversion commit and contain all necessary git-svn
+marks (so, you can recreate git-svn metadata locally). Note that right now
+mirrors reflect only <tt class="docutils literal"><span class="pre">trunk</span></tt> for each project. You can do the read-only Git
+clone of LLVM via:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git clone http://llvm.org/git/llvm.git
+</pre></div>
+</div>
+<p>If you want to check out clang too, run:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nb">cd </span>llvm/tools
+<span class="gp">%</span> git clone http://llvm.org/git/clang.git
+</pre></div>
+</div>
+<p>If you want to check out compiler-rt too, run:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nb">cd </span>llvm/projects
+<span class="gp">%</span> git clone http://llvm.org/git/compiler-rt.git
+</pre></div>
+</div>
+<p>If you want to check out the Test Suite Source Code (optional), run:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nb">cd </span>llvm/projects
+<span class="gp">%</span> git clone http://llvm.org/git/test-suite.git
+</pre></div>
+</div>
+<p>Since the upstream repository is in Subversion, you should use <tt class="docutils literal"><span class="pre">git</span>
+<span class="pre">pull</span> <span class="pre">--rebase</span></tt> instead of <tt class="docutils literal"><span class="pre">git</span> <span class="pre">pull</span></tt> to avoid generating a non-linear history
+in your clone.  To configure <tt class="docutils literal"><span class="pre">git</span> <span class="pre">pull</span></tt> to pass <tt class="docutils literal"><span class="pre">--rebase</span></tt> by default on the
+master branch, run the following command:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git config branch.master.rebase <span class="nb">true</span>
+</pre></div>
+</div>
+<div class="section" id="sending-patches-with-git">
+<h4><a class="toc-backref" href="#id18">Sending patches with Git</a><a class="headerlink" href="#sending-patches-with-git" title="Permalink to this headline">¶</a></h4>
+<p>Please read <a class="reference external" href="DeveloperPolicy.html#one-off-patches">Developer Policy</a>, too.</p>
+<p>Assume <tt class="docutils literal"><span class="pre">master</span></tt> points the upstream and <tt class="docutils literal"><span class="pre">mybranch</span></tt> points your working
+branch, and <tt class="docutils literal"><span class="pre">mybranch</span></tt> is rebased onto <tt class="docutils literal"><span class="pre">master</span></tt>.  At first you may check
+sanity of whitespaces:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git diff --check master..mybranch
+</pre></div>
+</div>
+<p>The easiest way to generate a patch is as below:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git diff master..mybranch > /path/to/mybranch.diff
+</pre></div>
+</div>
+<p>It is a little different from svn-generated diff. git-diff-generated diff has
+prefixes like <tt class="docutils literal"><span class="pre">a/</span></tt> and <tt class="docutils literal"><span class="pre">b/</span></tt>. Don’t worry, most developers might know it
+could be accepted with <tt class="docutils literal"><span class="pre">patch</span> <span class="pre">-p1</span> <span class="pre">-N</span></tt>.</p>
+<p>But you may generate patchset with git-format-patch. It generates by-each-commit
+patchset. To generate patch files to attach to your article:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git format-patch --no-attach master..mybranch -o /path/to/your/patchset
+</pre></div>
+</div>
+<p>If you would like to send patches directly, you may use git-send-email or
+git-imap-send. Here is an example to generate the patchset in Gmail’s [Drafts].</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git format-patch --attach master..mybranch --stdout | git imap-send
+</pre></div>
+</div>
+<p>Then, your .git/config should have [imap] sections.</p>
+<div class="highlight-ini"><div class="highlight"><pre><span class="k">[imap]</span>
+      <span class="na">host</span> <span class="o">=</span> <span class="s">imaps://imap.gmail.com</span>
+<span class="s">      user = your.gmail.account at gmail.com</span>
+<span class="s">      pass = himitsu!</span>
+<span class="s">      port = 993</span>
+<span class="s">      sslverify = false</span>
+<span class="c1">; in English</span>
+      <span class="na">folder</span> <span class="o">=</span> <span class="s">"[Gmail]/Drafts"</span>
+<span class="c1">; example for Japanese, "Modified UTF-7" encoded.</span>
+      <span class="na">folder</span> <span class="o">=</span> <span class="s">"[Gmail]/&Tgtm+DBN-"</span>
+<span class="c1">; example for Traditional Chinese</span>
+      <span class="na">folder</span> <span class="o">=</span> <span class="s">"[Gmail]/&g0l6Pw-"</span>
+</pre></div>
+</div>
+</div>
+<div class="section" id="for-developers-to-work-with-git-svn">
+<h4><a class="toc-backref" href="#id19">For developers to work with git-svn</a><a class="headerlink" href="#for-developers-to-work-with-git-svn" title="Permalink to this headline">¶</a></h4>
+<p>To set up clone from which you can submit code using <tt class="docutils literal"><span class="pre">git-svn</span></tt>, run:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git clone http://llvm.org/git/llvm.git
+<span class="gp">%</span> <span class="nb">cd </span>llvm
+<span class="gp">%</span> git svn init https://llvm.org/svn/llvm-project/llvm/trunk --username<span class="o">=</span><username>
+<span class="gp">%</span> git config svn-remote.svn.fetch :refs/remotes/origin/master
+<span class="gp">%</span> git svn rebase -l  <span class="c"># -l avoids fetching ahead of the git mirror.</span>
+
+<span class="gp">#</span> If you have clang too:
+<span class="gp">%</span> <span class="nb">cd </span>tools
+<span class="gp">%</span> git clone http://llvm.org/git/clang.git
+<span class="gp">%</span> <span class="nb">cd </span>clang
+<span class="gp">%</span> git svn init https://llvm.org/svn/llvm-project/cfe/trunk --username<span class="o">=</span><username>
+<span class="gp">%</span> git config svn-remote.svn.fetch :refs/remotes/origin/master
+<span class="gp">%</span> git svn rebase -l
+</pre></div>
+</div>
+<p>Likewise for compiler-rt and test-suite.</p>
+<p>To update this clone without generating git-svn tags that conflict with the
+upstream Git repo, run:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git fetch <span class="o">&&</span> <span class="o">(</span><span class="nb">cd </span>tools/clang <span class="o">&&</span> git fetch<span class="o">)</span>  <span class="c"># Get matching revisions of both trees.</span>
+<span class="gp">%</span> git checkout master
+<span class="gp">%</span> git svn rebase -l
+<span class="gp">%</span> <span class="o">(</span><span class="nb">cd </span>tools/clang <span class="o">&&</span>
+<span class="go">   git checkout master &&</span>
+<span class="go">   git svn rebase -l)</span>
+</pre></div>
+</div>
+<p>Likewise for compiler-rt and test-suite.</p>
+<p>This leaves your working directories on their master branches, so you’ll need to
+<tt class="docutils literal"><span class="pre">checkout</span></tt> each working branch individually and <tt class="docutils literal"><span class="pre">rebase</span></tt> it on top of its
+parent branch.</p>
+<p>For those who wish to be able to update an llvm repo/revert patches easily using
+git-svn, please look in the directory for the scripts <tt class="docutils literal"><span class="pre">git-svnup</span></tt> and
+<tt class="docutils literal"><span class="pre">git-svnrevert</span></tt>.</p>
+<p>To perform the aforementioned update steps go into your source directory and
+just type <tt class="docutils literal"><span class="pre">git-svnup</span></tt> or <tt class="docutils literal"><span class="pre">git</span> <span class="pre">svnup</span></tt> and everything will just work.</p>
+<p>If one wishes to revert a commit with git-svn, but do not want the git hash to
+escape into the commit message, one can use the script <tt class="docutils literal"><span class="pre">git-svnrevert</span></tt> or
+<tt class="docutils literal"><span class="pre">git</span> <span class="pre">svnrevert</span></tt> which will take in the git hash for the commit you want to
+revert, look up the appropriate svn revision, and output a message where all
+references to the git hash have been replaced with the svn revision.</p>
+<p>To commit back changes via git-svn, use <tt class="docutils literal"><span class="pre">git</span> <span class="pre">svn</span> <span class="pre">dcommit</span></tt>:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> git svn dcommit
+</pre></div>
+</div>
+<p>Note that git-svn will create one SVN commit for each Git commit you have pending,
+so squash and edit each commit before executing <tt class="docutils literal"><span class="pre">dcommit</span></tt> to make sure they all
+conform to the coding standards and the developers’ policy.</p>
+<p>On success, <tt class="docutils literal"><span class="pre">dcommit</span></tt> will rebase against the HEAD of SVN, so to avoid conflict,
+please make sure your current branch is up-to-date (via fetch/rebase) before
+proceeding.</p>
+<p>The git-svn metadata can get out of sync after you mess around with branches and
+<tt class="docutils literal"><span class="pre">dcommit</span></tt>. When that happens, <tt class="docutils literal"><span class="pre">git</span> <span class="pre">svn</span> <span class="pre">dcommit</span></tt> stops working, complaining
+about files with uncommitted changes. The fix is to rebuild the metadata:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> rm -rf .git/svn
+<span class="gp">%</span> git svn rebase -l
+</pre></div>
+</div>
+<p>Please, refer to the Git-SVN manual (<tt class="docutils literal"><span class="pre">man</span> <span class="pre">git-svn</span></tt>) for more information.</p>
+</div>
+</div>
+<div class="section" id="local-llvm-configuration">
+<h3><a class="toc-backref" href="#id20">Local LLVM Configuration</a><a class="headerlink" href="#local-llvm-configuration" title="Permalink to this headline">¶</a></h3>
+<p>Once checked out from the Subversion repository, the LLVM suite source code must
+be configured via the <tt class="docutils literal"><span class="pre">configure</span></tt> script.  This script sets variables in the
+various <tt class="docutils literal"><span class="pre">*.in</span></tt> files, most notably <tt class="docutils literal"><span class="pre">llvm/Makefile.config</span></tt> and
+<tt class="docutils literal"><span class="pre">llvm/include/Config/config.h</span></tt>.  It also populates <em>OBJ_ROOT</em> with the
+Makefiles needed to begin building LLVM.</p>
+<p>The following environment variables are used by the <tt class="docutils literal"><span class="pre">configure</span></tt> script to
+configure the build system:</p>
+<table border="1" class="docutils">
+<colgroup>
+<col width="17%" />
+<col width="83%" />
+</colgroup>
+<thead valign="bottom">
+<tr class="row-odd"><th class="head">Variable</th>
+<th class="head">Purpose</th>
+</tr>
+</thead>
+<tbody valign="top">
+<tr class="row-even"><td>CC</td>
+<td>Tells <tt class="docutils literal"><span class="pre">configure</span></tt> which C compiler to use.  By default,
+<tt class="docutils literal"><span class="pre">configure</span></tt> will check <tt class="docutils literal"><span class="pre">PATH</span></tt> for <tt class="docutils literal"><span class="pre">clang</span></tt> and GCC C
+compilers (in this order).  Use this variable to override
+<tt class="docutils literal"><span class="pre">configure</span></tt>‘s  default behavior.</td>
+</tr>
+<tr class="row-odd"><td>CXX</td>
+<td>Tells <tt class="docutils literal"><span class="pre">configure</span></tt> which C++ compiler to use.  By
+default, <tt class="docutils literal"><span class="pre">configure</span></tt> will check <tt class="docutils literal"><span class="pre">PATH</span></tt> for
+<tt class="docutils literal"><span class="pre">clang++</span></tt> and GCC C++ compilers (in this order).  Use
+this variable to override  <tt class="docutils literal"><span class="pre">configure</span></tt>‘s default
+behavior.</td>
+</tr>
+</tbody>
+</table>
+<p>The following options can be used to set or enable LLVM specific options:</p>
+<p><tt class="docutils literal"><span class="pre">--enable-optimized</span></tt></p>
+<blockquote>
+<div>Enables optimized compilation (debugging symbols are removed and GCC
+optimization flags are enabled). Note that this is the default setting if you
+are using the LLVM distribution. The default behavior of a Subversion
+checkout is to use an unoptimized build (also known as a debug build).</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">--enable-debug-runtime</span></tt></p>
+<blockquote>
+<div>Enables debug symbols in the runtime libraries. The default is to strip debug
+symbols from the runtime libraries.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">--enable-jit</span></tt></p>
+<blockquote>
+<div>Compile the Just In Time (JIT) compiler functionality.  This is not available
+on all platforms.  The default is dependent on platform, so it is best to
+explicitly enable it if you want it.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">--enable-targets=target-option</span></tt></p>
+<blockquote>
+<div><p>Controls which targets will be built and linked into llc. The default value
+for <tt class="docutils literal"><span class="pre">target_options</span></tt> is “all” which builds and links all available targets.
+The “host” target is selected as the target of the build host. You can also
+specify a comma separated list of target names that you want available in llc.
+The target names use all lower case. The current set of targets is:</p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">aarch64,</span> <span class="pre">arm,</span> <span class="pre">arm64,</span> <span class="pre">cpp,</span> <span class="pre">hexagon,</span> <span class="pre">mips,</span> <span class="pre">mipsel,</span> <span class="pre">mips64,</span> <span class="pre">mips64el,</span> <span class="pre">msp430,</span>
+<span class="pre">powerpc,</span> <span class="pre">nvptx,</span> <span class="pre">r600,</span> <span class="pre">sparc,</span> <span class="pre">systemz,</span> <span class="pre">x86,</span> <span class="pre">x86_64,</span> <span class="pre">xcore</span></tt>.</div></blockquote>
+</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">--enable-doxygen</span></tt></p>
+<blockquote>
+<div>Look for the doxygen program and enable construction of doxygen based
+documentation from the source code. This is disabled by default because
+generating the documentation can take a long time and producess 100s of
+megabytes of output.</div></blockquote>
+<p>To configure LLVM, follow these steps:</p>
+<ol class="arabic">
+<li><p class="first">Change directory into the object root directory:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nb">cd </span>OBJ_ROOT
+</pre></div>
+</div>
+</li>
+<li><p class="first">Run the <tt class="docutils literal"><span class="pre">configure</span></tt> script located in the LLVM source tree:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> SRC_ROOT/configure --prefix<span class="o">=</span>/install/path <span class="o">[</span>other options<span class="o">]</span>
+</pre></div>
+</div>
+</li>
+</ol>
+</div>
+<div class="section" id="compiling-the-llvm-suite-source-code">
+<h3><a class="toc-backref" href="#id21">Compiling the LLVM Suite Source Code</a><a class="headerlink" href="#compiling-the-llvm-suite-source-code" title="Permalink to this headline">¶</a></h3>
+<p>Once you have configured LLVM, you can build it.  There are three types of
+builds:</p>
+<p>Debug Builds</p>
+<blockquote>
+<div>These builds are the default when one is using a Subversion checkout and
+types <tt class="docutils literal"><span class="pre">gmake</span></tt> (unless the <tt class="docutils literal"><span class="pre">--enable-optimized</span></tt> option was used during
+configuration).  The build system will compile the tools and libraries with
+debugging information.  To get a Debug Build using the LLVM distribution the
+<tt class="docutils literal"><span class="pre">--disable-optimized</span></tt> option must be passed to <tt class="docutils literal"><span class="pre">configure</span></tt>.</div></blockquote>
+<p>Release (Optimized) Builds</p>
+<blockquote>
+<div>These builds are enabled with the <tt class="docutils literal"><span class="pre">--enable-optimized</span></tt> option to
+<tt class="docutils literal"><span class="pre">configure</span></tt> or by specifying <tt class="docutils literal"><span class="pre">ENABLE_OPTIMIZED=1</span></tt> on the <tt class="docutils literal"><span class="pre">gmake</span></tt> command
+line.  For these builds, the build system will compile the tools and libraries
+with GCC optimizations enabled and strip debugging information from the
+libraries and executables it generates.  Note that Release Builds are default
+when using an LLVM distribution.</div></blockquote>
+<p>Profile Builds</p>
+<blockquote>
+<div>These builds are for use with profiling.  They compile profiling information
+into the code for use with programs like <tt class="docutils literal"><span class="pre">gprof</span></tt>.  Profile builds must be
+started by specifying <tt class="docutils literal"><span class="pre">ENABLE_PROFILING=1</span></tt> on the <tt class="docutils literal"><span class="pre">gmake</span></tt> command line.</div></blockquote>
+<p>Once you have LLVM configured, you can build it by entering the <em>OBJ_ROOT</em>
+directory and issuing the following command:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> gmake
+</pre></div>
+</div>
+<p>If the build fails, please <a class="reference internal" href="#check-here">check here</a> to see if you are using a version of
+GCC that is known not to compile LLVM.</p>
+<p>If you have multiple processors in your machine, you may wish to use some of the
+parallel build options provided by GNU Make.  For example, you could use the
+command:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> gmake -j2
+</pre></div>
+</div>
+<p>There are several special targets which are useful when working with the LLVM
+source code:</p>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">clean</span></tt></p>
+<blockquote>
+<div>Removes all files generated by the build.  This includes object files,
+generated C/C++ files, libraries, and executables.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">dist-clean</span></tt></p>
+<blockquote>
+<div>Removes everything that <tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">clean</span></tt> does, but also removes files generated
+by <tt class="docutils literal"><span class="pre">configure</span></tt>.  It attempts to return the source tree to the original state
+in which it was shipped.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">install</span></tt></p>
+<blockquote>
+<div>Installs LLVM header files, libraries, tools, and documentation in a hierarchy
+under <tt class="docutils literal"><span class="pre">$PREFIX</span></tt>, specified with <tt class="docutils literal"><span class="pre">./configure</span> <span class="pre">--prefix=[dir]</span></tt>, which
+defaults to <tt class="docutils literal"><span class="pre">/usr/local</span></tt>.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">-C</span> <span class="pre">runtime</span> <span class="pre">install-bytecode</span></tt></p>
+<blockquote>
+<div>Assuming you built LLVM into $OBJDIR, when this command is run, it will
+install bitcode libraries into the GCC front end’s bitcode library directory.
+If you need to update your bitcode libraries, this is the target to use once
+you’ve built them.</div></blockquote>
+<p>Please see the <a class="reference external" href="MakefileGuide.html">Makefile Guide</a> for further details on
+these <tt class="docutils literal"><span class="pre">make</span></tt> targets and descriptions of other targets available.</p>
+<p>It is also possible to override default values from <tt class="docutils literal"><span class="pre">configure</span></tt> by declaring
+variables on the command line.  The following are some examples:</p>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">ENABLE_OPTIMIZED=1</span></tt></p>
+<blockquote>
+<div>Perform a Release (Optimized) build.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">ENABLE_OPTIMIZED=1</span> <span class="pre">DISABLE_ASSERTIONS=1</span></tt></p>
+<blockquote>
+<div>Perform a Release (Optimized) build without assertions enabled.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">ENABLE_OPTIMIZED=0</span></tt></p>
+<blockquote>
+<div>Perform a Debug build.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">ENABLE_PROFILING=1</span></tt></p>
+<blockquote>
+<div>Perform a Profiling build.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">VERBOSE=1</span></tt></p>
+<blockquote>
+<div>Print what <tt class="docutils literal"><span class="pre">gmake</span></tt> is doing on standard output.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">gmake</span> <span class="pre">TOOL_VERBOSE=1</span></tt></p>
+<blockquote>
+<div>Ask each tool invoked by the makefiles to print out what it is doing on
+the standard output. This also implies <tt class="docutils literal"><span class="pre">VERBOSE=1</span></tt>.</div></blockquote>
+<p>Every directory in the LLVM object tree includes a <tt class="docutils literal"><span class="pre">Makefile</span></tt> to build it and
+any subdirectories that it contains.  Entering any directory inside the LLVM
+object tree and typing <tt class="docutils literal"><span class="pre">gmake</span></tt> should rebuild anything in or below that
+directory that is out of date.</p>
+<p>This does not apply to building the documentation.
+LLVM’s (non-Doxygen) documentation is produced with the
+<a class="reference external" href="http://sphinx-doc.org/">Sphinx</a> documentation generation system.
+There are some HTML documents that have not yet been converted to the new
+system (which uses the easy-to-read and easy-to-write
+<a class="reference external" href="http://sphinx-doc.org/rest.html">reStructuredText</a> plaintext markup
+language).
+The generated documentation is built in the <tt class="docutils literal"><span class="pre">SRC_ROOT/docs</span></tt> directory using
+a special makefile.
+For instructions on how to install Sphinx, see
+<a class="reference external" href="http://lld.llvm.org/sphinx_intro.html">Sphinx Introduction for LLVM Developers</a>.
+After following the instructions there for installing Sphinx, build the LLVM
+HTML documentation by doing the following:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">$</span> <span class="nb">cd </span>SRC_ROOT/docs
+<span class="gp">$</span> make -f Makefile.sphinx
+</pre></div>
+</div>
+<p>This creates a <tt class="docutils literal"><span class="pre">_build/html</span></tt> sub-directory with all of the HTML files, not
+just the generated ones.
+This directory corresponds to <tt class="docutils literal"><span class="pre">llvm.org/docs</span></tt>.
+For example, <tt class="docutils literal"><span class="pre">_build/html/SphinxQuickstartTemplate.html</span></tt> corresponds to
+<tt class="docutils literal"><span class="pre">llvm.org/docs/SphinxQuickstartTemplate.html</span></tt>.
+The <a class="reference internal" href="SphinxQuickstartTemplate.html"><em>Sphinx Quickstart Template</em></a> is useful when creating a new document.</p>
+</div>
+<div class="section" id="cross-compiling-llvm">
+<h3><a class="toc-backref" href="#id22">Cross-Compiling LLVM</a><a class="headerlink" href="#cross-compiling-llvm" title="Permalink to this headline">¶</a></h3>
+<p>It is possible to cross-compile LLVM itself. That is, you can create LLVM
+executables and libraries to be hosted on a platform different from the platform
+where they are built (a Canadian Cross build). To configure a cross-compile,
+supply the configure script with <tt class="docutils literal"><span class="pre">--build</span></tt> and <tt class="docutils literal"><span class="pre">--host</span></tt> options that are
+different. The values of these options must be legal target triples that your
+GCC compiler supports.</p>
+<p>The result of such a build is executables that are not runnable on on the build
+host (–build option) but can be executed on the compile host (–host option).</p>
+<p>Check <a class="reference internal" href="HowToCrossCompileLLVM.html"><em>How To Cross-Compile Clang/LLVM using Clang/LLVM</em></a> and <a class="reference external" href="http://clang.llvm.org/docs/CrossCompilation.html">Clang docs on how to cross-compile in general</a> for more information
+about cross-compiling.</p>
+</div>
+<div class="section" id="the-location-of-llvm-object-files">
+<h3><a class="toc-backref" href="#id23">The Location of LLVM Object Files</a><a class="headerlink" href="#the-location-of-llvm-object-files" title="Permalink to this headline">¶</a></h3>
+<p>The LLVM build system is capable of sharing a single LLVM source tree among
+several LLVM builds.  Hence, it is possible to build LLVM for several different
+platforms or configurations using the same source tree.</p>
+<p>This is accomplished in the typical autoconf manner:</p>
+<ul>
+<li><p class="first">Change directory to where the LLVM object files should live:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> <span class="nb">cd </span>OBJ_ROOT
+</pre></div>
+</div>
+</li>
+<li><p class="first">Run the <tt class="docutils literal"><span class="pre">configure</span></tt> script found in the LLVM source directory:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> SRC_ROOT/configure
+</pre></div>
+</div>
+</li>
+</ul>
+<p>The LLVM build will place files underneath <em>OBJ_ROOT</em> in directories named after
+the build type:</p>
+<p>Debug Builds with assertions enabled (the default)</p>
+<blockquote>
+<div><p>Tools</p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">OBJ_ROOT/Debug+Asserts/bin</span></tt></div></blockquote>
+<p>Libraries</p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">OBJ_ROOT/Debug+Asserts/lib</span></tt></div></blockquote>
+</div></blockquote>
+<p>Release Builds</p>
+<blockquote>
+<div><p>Tools</p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">OBJ_ROOT/Release/bin</span></tt></div></blockquote>
+<p>Libraries</p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">OBJ_ROOT/Release/lib</span></tt></div></blockquote>
+</div></blockquote>
+<p>Profile Builds</p>
+<blockquote>
+<div><p>Tools</p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">OBJ_ROOT/Profile/bin</span></tt></div></blockquote>
+<p>Libraries</p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">OBJ_ROOT/Profile/lib</span></tt></div></blockquote>
+</div></blockquote>
+</div>
+<div class="section" id="optional-configuration-items">
+<h3><a class="toc-backref" href="#id24">Optional Configuration Items</a><a class="headerlink" href="#optional-configuration-items" title="Permalink to this headline">¶</a></h3>
+<p>If you’re running on a Linux system that supports the <a class="reference external" href="http://en.wikipedia.org/wiki/binfmt_misc">binfmt_misc</a>
+module, and you have root access on the system, you can set your system up to
+execute LLVM bitcode files directly. To do this, use commands like this (the
+first command may not be required if you are already using the module):</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> mount -t binfmt_misc none /proc/sys/fs/binfmt_misc
+<span class="gp">%</span> <span class="nb">echo</span> <span class="s1">':llvm:M::BC::/path/to/lli:'</span> > /proc/sys/fs/binfmt_misc/register
+<span class="gp">%</span> chmod u+x hello.bc   <span class="o">(</span><span class="k">if </span>needed<span class="o">)</span>
+<span class="gp">%</span> ./hello.bc
+</pre></div>
+</div>
+<p>This allows you to execute LLVM bitcode files directly.  On Debian, you can also
+use this command instead of the ‘echo’ command above:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> sudo update-binfmts --install llvm /path/to/lli --magic <span class="s1">'BC'</span>
+</pre></div>
+</div>
+</div>
+</div>
+<div class="section" id="general-layout">
+<span id="program-layout"></span><span id="id3"></span><h2><a class="toc-backref" href="#id25">Program Layout</a><a class="headerlink" href="#general-layout" title="Permalink to this headline">¶</a></h2>
+<p>One useful source of information about the LLVM source base is the LLVM <a class="reference external" href="http://www.doxygen.org/">doxygen</a> documentation available at
+<a class="reference external" href="http://llvm.org/doxygen/">http://llvm.org/doxygen/</a>.  The following is a brief introduction to code
+layout:</p>
+<div class="section" id="llvm-examples">
+<h3><a class="toc-backref" href="#id26"><tt class="docutils literal"><span class="pre">llvm/examples</span></tt></a><a class="headerlink" href="#llvm-examples" title="Permalink to this headline">¶</a></h3>
+<p>This directory contains some simple examples of how to use the LLVM IR and JIT.</p>
+</div>
+<div class="section" id="llvm-include">
+<h3><a class="toc-backref" href="#id27"><tt class="docutils literal"><span class="pre">llvm/include</span></tt></a><a class="headerlink" href="#llvm-include" title="Permalink to this headline">¶</a></h3>
+<p>This directory contains public header files exported from the LLVM library. The
+three main subdirectories of this directory are:</p>
+<p><tt class="docutils literal"><span class="pre">llvm/include/llvm</span></tt></p>
+<blockquote>
+<div>This directory contains all of the LLVM specific header files.  This directory
+also has subdirectories for different portions of LLVM: <tt class="docutils literal"><span class="pre">Analysis</span></tt>,
+<tt class="docutils literal"><span class="pre">CodeGen</span></tt>, <tt class="docutils literal"><span class="pre">Target</span></tt>, <tt class="docutils literal"><span class="pre">Transforms</span></tt>, etc...</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/include/llvm/Support</span></tt></p>
+<blockquote>
+<div>This directory contains generic support libraries that are provided with LLVM
+but not necessarily specific to LLVM. For example, some C++ STL utilities and
+a Command Line option processing library store their header files here.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/include/llvm/Config</span></tt></p>
+<blockquote>
+<div>This directory contains header files configured by the <tt class="docutils literal"><span class="pre">configure</span></tt> script.
+They wrap “standard” UNIX and C header files.  Source code can include these
+header files which automatically take care of the conditional #includes that
+the <tt class="docutils literal"><span class="pre">configure</span></tt> script generates.</div></blockquote>
+</div>
+<div class="section" id="llvm-lib">
+<h3><a class="toc-backref" href="#id28"><tt class="docutils literal"><span class="pre">llvm/lib</span></tt></a><a class="headerlink" href="#llvm-lib" title="Permalink to this headline">¶</a></h3>
+<p>This directory contains most of the source files of the LLVM system. In LLVM,
+almost all code exists in libraries, making it very easy to share code among the
+different <a class="reference internal" href="#tools">tools</a>.</p>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/VMCore/</span></tt></p>
+<blockquote>
+<div>This directory holds the core LLVM source files that implement core classes
+like Instruction and BasicBlock.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/AsmParser/</span></tt></p>
+<blockquote>
+<div>This directory holds the source code for the LLVM assembly language parser
+library.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/Bitcode/</span></tt></p>
+<blockquote>
+<div>This directory holds code for reading and write LLVM bitcode.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/Analysis/</span></tt></p>
+<blockquote>
+<div>This directory contains a variety of different program analyses, such as
+Dominator Information, Call Graphs, Induction Variables, Interval
+Identification, Natural Loop Identification, etc.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/Transforms/</span></tt></p>
+<blockquote>
+<div>This directory contains the source code for the LLVM to LLVM program
+transformations, such as Aggressive Dead Code Elimination, Sparse Conditional
+Constant Propagation, Inlining, Loop Invariant Code Motion, Dead Global
+Elimination, and many others.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/Target/</span></tt></p>
+<blockquote>
+<div>This directory contains files that describe various target architectures for
+code generation.  For example, the <tt class="docutils literal"><span class="pre">llvm/lib/Target/X86</span></tt> directory holds the
+X86 machine description while <tt class="docutils literal"><span class="pre">llvm/lib/Target/ARM</span></tt> implements the ARM
+backend.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/CodeGen/</span></tt></p>
+<blockquote>
+<div>This directory contains the major parts of the code generator: Instruction
+Selector, Instruction Scheduling, and Register Allocation.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/MC/</span></tt></p>
+<blockquote>
+<div>(FIXME: T.B.D.)</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/Debugger/</span></tt></p>
+<blockquote>
+<div>This directory contains the source level debugger library that makes it
+possible to instrument LLVM programs so that a debugger could identify source
+code locations at which the program is executing.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/ExecutionEngine/</span></tt></p>
+<blockquote>
+<div>This directory contains libraries for executing LLVM bitcode directly at
+runtime in both interpreted and JIT compiled fashions.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm/lib/Support/</span></tt></p>
+<blockquote>
+<div>This directory contains the source code that corresponds to the header files
+located in <tt class="docutils literal"><span class="pre">llvm/include/ADT/</span></tt> and <tt class="docutils literal"><span class="pre">llvm/include/Support/</span></tt>.</div></blockquote>
+</div>
+<div class="section" id="llvm-projects">
+<h3><a class="toc-backref" href="#id29"><tt class="docutils literal"><span class="pre">llvm/projects</span></tt></a><a class="headerlink" href="#llvm-projects" title="Permalink to this headline">¶</a></h3>
+<p>This directory contains projects that are not strictly part of LLVM but are
+shipped with LLVM. This is also the directory where you should create your own
+LLVM-based projects.</p>
+</div>
+<div class="section" id="llvm-runtime">
+<h3><a class="toc-backref" href="#id30"><tt class="docutils literal"><span class="pre">llvm/runtime</span></tt></a><a class="headerlink" href="#llvm-runtime" title="Permalink to this headline">¶</a></h3>
+<p>This directory contains libraries which are compiled into LLVM bitcode and used
+when linking programs with the Clang front end.  Most of these libraries are
+skeleton versions of real libraries; for example, libc is a stripped down
+version of glibc.</p>
+<p>Unlike the rest of the LLVM suite, this directory needs the LLVM GCC front end
+to compile.</p>
+</div>
+<div class="section" id="llvm-test">
+<h3><a class="toc-backref" href="#id31"><tt class="docutils literal"><span class="pre">llvm/test</span></tt></a><a class="headerlink" href="#llvm-test" title="Permalink to this headline">¶</a></h3>
+<p>This directory contains feature and regression tests and other basic sanity
+checks on the LLVM infrastructure. These are intended to run quickly and cover a
+lot of territory without being exhaustive.</p>
+</div>
+<div class="section" id="test-suite">
+<h3><a class="toc-backref" href="#id32"><tt class="docutils literal"><span class="pre">test-suite</span></tt></a><a class="headerlink" href="#test-suite" title="Permalink to this headline">¶</a></h3>
+<p>This is not a directory in the normal llvm module; it is a separate Subversion
+module that must be checked out (usually to <tt class="docutils literal"><span class="pre">projects/test-suite</span></tt>).  This
+module contains a comprehensive correctness, performance, and benchmarking test
+suite for LLVM. It is a separate Subversion module because not every LLVM user
+is interested in downloading or building such a comprehensive test suite. For
+further details on this test suite, please see the <a class="reference internal" href="TestingGuide.html"><em>Testing Guide</em></a> document.</p>
+</div>
+<div class="section" id="llvm-tools">
+<span id="tools"></span><h3><a class="toc-backref" href="#id33"><tt class="docutils literal"><span class="pre">llvm/tools</span></tt></a><a class="headerlink" href="#llvm-tools" title="Permalink to this headline">¶</a></h3>
+<p>The <strong>tools</strong> directory contains the executables built out of the libraries
+above, which form the main part of the user interface.  You can always get help
+for a tool by typing <tt class="docutils literal"><span class="pre">tool_name</span> <span class="pre">-help</span></tt>.  The following is a brief introduction
+to the most important tools.  More detailed information is in
+the <a class="reference external" href="CommandGuide/index.html">Command Guide</a>.</p>
+<p><tt class="docutils literal"><span class="pre">bugpoint</span></tt></p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">bugpoint</span></tt> is used to debug optimization passes or code generation backends
+by narrowing down the given test case to the minimum number of passes and/or
+instructions that still cause a problem, whether it is a crash or
+miscompilation. See <a class="reference external" href="HowToSubmitABug.html">HowToSubmitABug.html</a> for more information on using
+<tt class="docutils literal"><span class="pre">bugpoint</span></tt>.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm-ar</span></tt></p>
+<blockquote>
+<div>The archiver produces an archive containing the given LLVM bitcode files,
+optionally with an index for faster lookup.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm-as</span></tt></p>
+<blockquote>
+<div>The assembler transforms the human readable LLVM assembly to LLVM bitcode.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm-dis</span></tt></p>
+<blockquote>
+<div>The disassembler transforms the LLVM bitcode to human readable LLVM assembly.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvm-link</span></tt></p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">llvm-link</span></tt>, not surprisingly, links multiple LLVM modules into a single
+program.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">lli</span></tt></p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">lli</span></tt> is the LLVM interpreter, which can directly execute LLVM bitcode
+(although very slowly...). For architectures that support it (currently x86,
+Sparc, and PowerPC), by default, <tt class="docutils literal"><span class="pre">lli</span></tt> will function as a Just-In-Time
+compiler (if the functionality was compiled in), and will execute the code
+<em>much</em> faster than the interpreter.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llc</span></tt></p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">llc</span></tt> is the LLVM backend compiler, which translates LLVM bitcode to a
+native code assembly file or to C code (with the <tt class="docutils literal"><span class="pre">-march=c</span></tt> option).</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">opt</span></tt></p>
+<blockquote>
+<div><p><tt class="docutils literal"><span class="pre">opt</span></tt> reads LLVM bitcode, applies a series of LLVM to LLVM transformations
+(which are specified on the command line), and then outputs the resultant
+bitcode.  The ‘<tt class="docutils literal"><span class="pre">opt</span> <span class="pre">-help</span></tt>‘ command is a good way to get a list of the
+program transformations available in LLVM.</p>
+<p><tt class="docutils literal"><span class="pre">opt</span></tt> can also be used to run a specific analysis on an input LLVM bitcode
+file and print out the results.  It is primarily useful for debugging
+analyses, or familiarizing yourself with what an analysis does.</p>
+</div></blockquote>
+</div>
+<div class="section" id="llvm-utils">
+<h3><a class="toc-backref" href="#id34"><tt class="docutils literal"><span class="pre">llvm/utils</span></tt></a><a class="headerlink" href="#llvm-utils" title="Permalink to this headline">¶</a></h3>
+<p>This directory contains utilities for working with LLVM source code, and some of
+the utilities are actually required as part of the build process because they
+are code generators for parts of LLVM infrastructure.</p>
+<p><tt class="docutils literal"><span class="pre">codegen-diff</span></tt></p>
+<blockquote>
+<div><tt class="docutils literal"><span class="pre">codegen-diff</span></tt> is a script that finds differences between code that LLC
+generates and code that LLI generates. This is a useful tool if you are
+debugging one of them, assuming that the other generates correct output. For
+the full user manual, run <tt class="docutils literal"><span class="pre">`perldoc</span> <span class="pre">codegen-diff'</span></tt>.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">emacs/</span></tt></p>
+<blockquote>
+<div>The <tt class="docutils literal"><span class="pre">emacs</span></tt> directory contains syntax-highlighting files which will work
+with Emacs and XEmacs editors, providing syntax highlighting support for LLVM
+assembly files and TableGen description files. For information on how to use
+the syntax files, consult the <tt class="docutils literal"><span class="pre">README</span></tt> file in that directory.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">getsrcs.sh</span></tt></p>
+<blockquote>
+<div>The <tt class="docutils literal"><span class="pre">getsrcs.sh</span></tt> script finds and outputs all non-generated source files,
+which is useful if one wishes to do a lot of development across directories
+and does not want to individually find each file. One way to use it is to run,
+for example: <tt class="docutils literal"><span class="pre">xemacs</span> <span class="pre">`utils/getsources.sh`</span></tt> from the top of your LLVM source
+tree.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">llvmgrep</span></tt></p>
+<blockquote>
+<div>This little tool performs an <tt class="docutils literal"><span class="pre">egrep</span> <span class="pre">-H</span> <span class="pre">-n</span></tt> on each source file in LLVM and
+passes to it a regular expression provided on <tt class="docutils literal"><span class="pre">llvmgrep</span></tt>‘s command
+line. This is a very efficient way of searching the source base for a
+particular regular expression.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">makellvm</span></tt></p>
+<blockquote>
+<div>The <tt class="docutils literal"><span class="pre">makellvm</span></tt> script compiles all files in the current directory and then
+compiles and links the tool that is the first argument. For example, assuming
+you are in the directory <tt class="docutils literal"><span class="pre">llvm/lib/Target/Sparc</span></tt>, if <tt class="docutils literal"><span class="pre">makellvm</span></tt> is in your
+path, simply running <tt class="docutils literal"><span class="pre">makellvm</span> <span class="pre">llc</span></tt> will make a build of the current
+directory, switch to directory <tt class="docutils literal"><span class="pre">llvm/tools/llc</span></tt> and build it, causing a
+re-linking of LLC.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">TableGen/</span></tt></p>
+<blockquote>
+<div>The <tt class="docutils literal"><span class="pre">TableGen</span></tt> directory contains the tool used to generate register
+descriptions, instruction set descriptions, and even assemblers from common
+TableGen description files.</div></blockquote>
+<p><tt class="docutils literal"><span class="pre">vim/</span></tt></p>
+<blockquote>
+<div>The <tt class="docutils literal"><span class="pre">vim</span></tt> directory contains syntax-highlighting files which will work with
+the VIM editor, providing syntax highlighting support for LLVM assembly files
+and TableGen description files. For information on how to use the syntax
+files, consult the <tt class="docutils literal"><span class="pre">README</span></tt> file in that directory.</div></blockquote>
+</div>
+</div>
+<div class="section" id="an-example-using-the-llvm-tool-chain">
+<span id="simple-example"></span><h2><a class="toc-backref" href="#id35">An Example Using the LLVM Tool Chain</a><a class="headerlink" href="#an-example-using-the-llvm-tool-chain" title="Permalink to this headline">¶</a></h2>
+<p>This section gives an example of using LLVM with the Clang front end.</p>
+<div class="section" id="example-with-clang">
+<h3><a class="toc-backref" href="#id36">Example with clang</a><a class="headerlink" href="#example-with-clang" title="Permalink to this headline">¶</a></h3>
+<ol class="arabic">
+<li><p class="first">First, create a simple C file, name it ‘hello.c’:</p>
+<div class="highlight-c"><div class="highlight"><pre><span class="cp">#include <stdio.h></span>
+
+<span class="kt">int</span> <span class="nf">main</span><span class="p">()</span> <span class="p">{</span>
+  <span class="n">printf</span><span class="p">(</span><span class="s">"hello world</span><span class="se">\n</span><span class="s">"</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>
+</li>
+<li><p class="first">Next, compile the C file into a native executable:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> clang hello.c -o hello
+</pre></div>
+</div>
+<div class="admonition note">
+<p class="first admonition-title">Note</p>
+<p class="last">Clang works just like GCC by default.  The standard -S and -c arguments
+work as usual (producing a native .s or .o file, respectively).</p>
+</div>
+</li>
+<li><p class="first">Next, compile the C file into an LLVM bitcode file:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> clang -O3 -emit-llvm hello.c -c -o hello.bc
+</pre></div>
+</div>
+<p>The -emit-llvm option can be used with the -S or -c options to emit an LLVM
+<tt class="docutils literal"><span class="pre">.ll</span></tt> or <tt class="docutils literal"><span class="pre">.bc</span></tt> file (respectively) for the code.  This allows you to use
+the <a class="reference external" href="CommandGuide/index.html">standard LLVM tools</a> on the bitcode file.</p>
+</li>
+<li><p class="first">Run the program in both forms. To run the program, use:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> ./hello
+</pre></div>
+</div>
+<p>and</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> lli hello.bc
+</pre></div>
+</div>
+<p>The second examples shows how to invoke the LLVM JIT, <a class="reference internal" href="CommandGuide/lli.html"><em>lli</em></a>.</p>
+</li>
+<li><p class="first">Use the <tt class="docutils literal"><span class="pre">llvm-dis</span></tt> utility to take a look at the LLVM assembly code:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> llvm-dis < hello.bc | less
+</pre></div>
+</div>
+</li>
+<li><p class="first">Compile the program to native assembly using the LLC code generator:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> llc hello.bc -o hello.s
+</pre></div>
+</div>
+</li>
+<li><p class="first">Assemble the native assembly language file into a program:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> /opt/SUNWspro/bin/cc -xarch<span class="o">=</span>v9 hello.s -o hello.native   <span class="c"># On Solaris</span>
+
+<span class="gp">%</span> gcc hello.s -o hello.native                              <span class="c"># On others</span>
+</pre></div>
+</div>
+</li>
+<li><p class="first">Execute the native code program:</p>
+<div class="highlight-console"><div class="highlight"><pre><span class="gp">%</span> ./hello.native
+</pre></div>
+</div>
+<p>Note that using clang to compile directly to native code (i.e. when the
+<tt class="docutils literal"><span class="pre">-emit-llvm</span></tt> option is not present) does steps 6/7/8 for you.</p>
+</li>
+</ol>
+</div>
+</div>
+<div class="section" id="common-problems">
+<h2><a class="toc-backref" href="#id37">Common Problems</a><a class="headerlink" href="#common-problems" title="Permalink to this headline">¶</a></h2>
+<p>If you are having problems building or using LLVM, or if you have any other
+general questions about LLVM, please consult the <a class="reference external" href="FAQ.html">Frequently Asked
+Questions</a> page.</p>
+</div>
+<div class="section" id="links">
+<span id="id4"></span><h2><a class="toc-backref" href="#id38">Links</a><a class="headerlink" href="#links" title="Permalink to this headline">¶</a></h2>
+<p>This document is just an <strong>introduction</strong> on how to use LLVM to do some simple
+things... there are many more interesting and complicated things that you can do
+that aren’t documented here (but we’ll gladly accept a patch if you want to
+write something up!).  For more information about LLVM, check out:</p>
+<ul class="simple">
+<li><a class="reference external" href="http://llvm.org/">LLVM Homepage</a></li>
+<li><a class="reference external" href="http://llvm.org/doxygen/">LLVM Doxygen Tree</a></li>
+<li><a class="reference external" href="http://llvm.org/docs/Projects.html">Starting a Project that Uses LLVM</a></li>
+</ul>
+</div>
+</div>
+
+
+          </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related">
+      <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="GettingStartedVS.html" title="Getting Started with the LLVM System using Microsoft Visual Studio"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="CommandGuide/llvm-readobj.html" title="llvm-readobj - LLVM Object Reader"
+             >previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+    <div class="footer">
+        © Copyright 2003-2014, LLVM Project.
+      Last updated on 2015-01-13.
+      Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
+    </div>
+  </body>
+</html>
\ No newline at end of file

Added: www-releases/trunk/3.5.1/docs/GettingStartedVS.html
URL: http://llvm.org/viewvc/llvm-project/www-releases/trunk/3.5.1/docs/GettingStartedVS.html?rev=225843&view=auto
==============================================================================
--- www-releases/trunk/3.5.1/docs/GettingStartedVS.html (added)
+++ www-releases/trunk/3.5.1/docs/GettingStartedVS.html Tue Jan 13 16:55:20 2015
@@ -0,0 +1,331 @@
+
+
+<!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="Content-Type" content="text/html; charset=utf-8" />
+    
+    <title>Getting Started with the LLVM System using Microsoft Visual Studio — LLVM 3.5 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">
+      var DOCUMENTATION_OPTIONS = {
+        URL_ROOT:    '',
+        VERSION:     '3.5',
+        COLLAPSE_INDEX: false,
+        FILE_SUFFIX: '.html',
+        HAS_SOURCE:  true
+      };
+    </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>
+    <link rel="top" title="LLVM 3.5 documentation" href="index.html" />
+    <link rel="next" title="Frequently Asked Questions (FAQ)" href="FAQ.html" />
+    <link rel="prev" title="Getting Started with the LLVM System" href="GettingStarted.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">
+      <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="FAQ.html" title="Frequently Asked Questions (FAQ)"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="GettingStarted.html" title="Getting Started with the LLVM System"
+             accesskey="P">previous</a> |</li>
+  <li><a href="http://llvm.org/">LLVM Home</a> | </li>
+  <li><a href="index.html">Documentation</a>»</li>
+ 
+      </ul>
+    </div>
+
+
+    <div class="document">
+      <div class="documentwrapper">
+          <div class="body">
+            
+  <div class="section" id="getting-started-with-the-llvm-system-using-microsoft-visual-studio">
+<h1>Getting Started with the LLVM System using Microsoft Visual Studio<a class="headerlink" href="#getting-started-with-the-llvm-system-using-microsoft-visual-studio" title="Permalink to this headline">¶</a></h1>
+<div class="contents local topic" id="contents">
+<ul class="simple">
+<li><a class="reference internal" href="#overview" id="id2">Overview</a></li>
+<li><a class="reference internal" href="#requirements" id="id3">Requirements</a><ul>
+<li><a class="reference internal" href="#hardware" id="id4">Hardware</a></li>
+<li><a class="reference internal" href="#software" id="id5">Software</a></li>
+</ul>
+</li>
+<li><a class="reference internal" href="#getting-started" id="id6">Getting Started</a></li>
+<li><a class="reference internal" href="#an-example-using-the-llvm-tool-chain" id="id7">An Example Using the LLVM Tool Chain</a></li>
+<li><a class="reference internal" href="#common-problems" id="id8">Common Problems</a></li>
+<li><a class="reference internal" href="#links" id="id9">Links</a></li>
+</ul>
+</div>
+<div class="section" id="overview">
+<h2><a class="toc-backref" href="#id2">Overview</a><a class="headerlink" href="#overview" title="Permalink to this headline">¶</a></h2>
+<p>Welcome to LLVM on Windows! This document only covers LLVM on Windows using
+Visual Studio, not mingw or cygwin. In order to get started, you first need to
+know some basic information.</p>
+<p>There are many different projects that compose LLVM. The first is the LLVM
+suite. This contains all of the tools, libraries, and header files needed to
+use LLVM. It contains an assembler, disassembler,
+bitcode analyzer and bitcode optimizer. It also contains a test suite that can
+be used to test the LLVM tools.</p>
+<p>Another useful project on Windows is <a class="reference external" href="http://clang.llvm.org/">Clang</a>.
+Clang is a C family ([Objective]C/C++) compiler. Clang mostly works on
+Windows, but does not currently understand all of the Microsoft extensions
+to C and C++. Because of this, clang cannot parse the C++ standard library
+included with Visual Studio, nor parts of the Windows Platform SDK. However,
+most standard C programs do compile. Clang can be used to emit bitcode,
+directly emit object files or even linked executables using Visual Studio’s
+<tt class="docutils literal"><span class="pre">link.exe</span></tt>.</p>
+<p>The large LLVM test suite cannot be run on the Visual Studio port at this
+time.</p>
+<p>Most of the tools build and work.  <tt class="docutils literal"><span class="pre">bugpoint</span></tt> does build, but does
+not work.</p>
+<p>Additional information about the LLVM directory structure and tool chain
+can be found on the main <a class="reference internal" href="GettingStarted.html"><em>Getting Started with the LLVM System</em></a> page.</p>
+</div>
+<div class="section" id="requirements">
+<h2><a class="toc-backref" href="#id3">Requirements</a><a class="headerlink" href="#requirements" title="Permalink to this headline">¶</a></h2>
+<p>Before you begin to use the LLVM system, review the requirements given
+below.  This may save you some trouble by knowing ahead of time what hardware
+and software you will need.</p>
+<div class="section" id="hardware">
+<h3><a class="toc-backref" href="#id4">Hardware</a><a class="headerlink" href="#hardware" title="Permalink to this headline">¶</a></h3>
+<p>Any system that can adequately run Visual Studio 2012 is fine. The LLVM
+source tree and object files, libraries and executables will consume
+approximately 3GB.</p>
+</div>
+<div class="section" id="software">
+<h3><a class="toc-backref" href="#id5">Software</a><a class="headerlink" href="#software" title="Permalink to this headline">¶</a></h3>
+<p>You will need Visual Studio 2012 or higher.</p>
+<p>You will also need the <a class="reference external" href="http://www.cmake.org/">CMake</a> build system since it
+generates the project files you will use to build with.</p>
+<p>If you would like to run the LLVM tests you will need <a class="reference external" href="http://www.python.org/">Python</a>. Versions 2.4-2.7 are known to work. You will need
+<a class="reference external" href="http://gnuwin32.sourceforge.net/">GnuWin32</a> tools, too.</p>
+<p>Do not install the LLVM directory tree into a path containing spaces (e.g.
+<tt class="docutils literal"><span class="pre">C:\Documents</span> <span class="pre">and</span> <span class="pre">Settings\...</span></tt>) as the configure step will fail.</p>
+</div>
+</div>
+<div class="section" id="getting-started">
+<h2><a class="toc-backref" href="#id6">Getting Started</a><a class="headerlink" href="#getting-started" title="Permalink to this headline">¶</a></h2>
+<p>Here’s the short story for getting up and running quickly with LLVM:</p>
+<ol class="arabic">
+<li><p class="first">Read the documentation.</p>
+</li>
+<li><p class="first">Seriously, read the documentation.</p>
+</li>
+<li><p class="first">Remember that you were warned twice about reading the documentation.</p>
+</li>
+<li><p class="first">Get the Source Code</p>
+<ul>
+<li><p class="first">With the distributed files:</p>
+<blockquote>
+<div><ol class="arabic simple">
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre"><where-you-want-llvm-to-live></span></tt></li>
+<li><tt class="docutils literal"><span class="pre">gunzip</span> <span class="pre">--stdout</span> <span class="pre">llvm-VERSION.tar.gz</span> <span class="pre">|</span> <span class="pre">tar</span> <span class="pre">-xvf</span> <span class="pre">-</span></tt>
+(<em>or use WinZip</em>)</li>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">llvm</span></tt></li>
+</ol>
+</div></blockquote>
+</li>
+<li><p class="first">With anonymous Subversion access:</p>
+<blockquote>
+<div><ol class="arabic simple">
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre"><where-you-want-llvm-to-live></span></tt></li>
+<li><tt class="docutils literal"><span class="pre">svn</span> <span class="pre">co</span> <span class="pre">http://llvm.org/svn/llvm-project/llvm/trunk</span> <span class="pre">llvm</span></tt></li>
+<li><tt class="docutils literal"><span class="pre">cd</span> <span class="pre">llvm</span></tt></li>
+</ol>
+</div></blockquote>
+</li>
+</ul>
+</li>
+<li><p class="first">Use <a class="reference external" href="http://www.cmake.org/">CMake</a> to generate up-to-date project files:</p>
+<ul class="simple">
+<li>Once CMake is installed then the simplest way is to just start the
+CMake GUI, select the directory where you have LLVM extracted to, and
+the default options should all be fine.  One option you may really
+want to change, regardless of anything else, might be the
+<tt class="docutils literal"><span class="pre">CMAKE_INSTALL_PREFIX</span></tt> setting to select a directory to INSTALL to
+once compiling is complete, although installation is not mandatory for
+using LLVM.  Another important option is <tt class="docutils literal"><span class="pre">LLVM_TARGETS_TO_BUILD</span></tt>,
+which controls the LLVM target architectures that are included on the
+build.</li>
+<li>See the <a class="reference internal" href="CMake.html"><em>LLVM CMake guide</em></a> for detailed information about
+how to configure the LLVM build.</li>
+<li>CMake generates project files for all build types. To select a specific
+build type, use the Configuration manager from the VS IDE or the
+<tt class="docutils literal"><span class="pre">/property:Configuration</span></tt> command line option when using MSBuild.</li>
+</ul>
+</li>
+<li><p class="first">Start Visual Studio</p>
+<ul class="simple">
+<li>In the directory you created the project files will have an <tt class="docutils literal"><span class="pre">llvm.sln</span></tt>
+file, just double-click on that to open Visual Studio.</li>
+</ul>
+</li>
+<li><p class="first">Build the LLVM Suite:</p>
+<ul class="simple">
+<li>The projects may still be built individually, but to build them all do
+not just select all of them in batch build (as some are meant as
+configuration projects), but rather select and build just the
+<tt class="docutils literal"><span class="pre">ALL_BUILD</span></tt> project to build everything, or the <tt class="docutils literal"><span class="pre">INSTALL</span></tt> project,
+which first builds the <tt class="docutils literal"><span class="pre">ALL_BUILD</span></tt> project, then installs the LLVM
+headers, libs, and other useful things to the directory set by the
+<tt class="docutils literal"><span class="pre">CMAKE_INSTALL_PREFIX</span></tt> setting when you first configured CMake.</li>
+<li>The Fibonacci project is a sample program that uses the JIT. Modify the
+project’s debugging properties to provide a numeric command line argument
+or run it from the command line.  The program will print the
+corresponding fibonacci value.</li>
+</ul>
+</li>
+<li><p class="first">Test LLVM in Visual Studio:</p>
+<ul class="simple">
+<li>If <tt class="docutils literal"><span class="pre">%PATH%</span></tt> does not contain GnuWin32, you may specify
+<tt class="docutils literal"><span class="pre">LLVM_LIT_TOOLS_DIR</span></tt> on CMake for the path to GnuWin32.</li>
+<li>You can run LLVM tests by merely building the project “check”. The test
+results will be shown in the VS output window.</li>
+</ul>
+</li>
+<li><p class="first">Test LLVM on the command line:</p>
+<ul>
+<li><p class="first">The LLVM tests can be run by changing directory to the llvm source
+directory and running:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..\llvm<span class="p">></span> <span class="n">python</span> ..\build\bin\llvm-lit --param build_config<span class="o">=</span>Win<span class="m">32</span> --param build_mode<span class="o">=</span>Debug --param llvm_site_config<span class="o">=</span>..<span class="n">/build/test/lit.site.cfg</span> test
+</pre></div>
+</div>
+<p>This example assumes that Python is in your PATH variable, you
+have built a Win32 Debug version of llvm with a standard out of
+line build. You should not see any unexpected failures, but will
+see many unsupported tests and expected failures.</p>
+<p>A specific test or test directory can be run with:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..\llvm<span class="p">></span> <span class="n">python</span> ..\build\bin\llvm-lit --param build_config<span class="o">=</span>Win<span class="m">32</span> --param build_mode<span class="o">=</span>Debug --param llvm_site_config<span class="o">=</span>..<span class="n">/build/test/lit.site.cfg</span> test<span class="n">/path/to/test</span>
+</pre></div>
+</div>
+</li>
+</ul>
+</li>
+</ol>
+</div>
+<div class="section" id="an-example-using-the-llvm-tool-chain">
+<h2><a class="toc-backref" href="#id7">An Example Using the LLVM Tool Chain</a><a class="headerlink" href="#an-example-using-the-llvm-tool-chain" title="Permalink to this headline">¶</a></h2>
+<ol class="arabic">
+<li><p class="first">First, create a simple C file, name it ‘<tt class="docutils literal"><span class="pre">hello.c</span></tt>‘:</p>
+<div class="highlight-c"><div class="highlight"><pre><span class="cp">#include <stdio.h></span>
+<span class="kt">int</span> <span class="nf">main</span><span class="p">()</span> <span class="p">{</span>
+  <span class="n">printf</span><span class="p">(</span><span class="s">"hello world</span><span class="se">\n</span><span class="s">"</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>
+</li>
+<li><p class="first">Next, compile the C file into an LLVM bitcode file:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..<span class="p">></span> <span class="n">clang</span> -c hello.c -emit-llvm -o hello.bc
+</pre></div>
+</div>
+<p>This will create the result file <tt class="docutils literal"><span class="pre">hello.bc</span></tt> which is the LLVM bitcode
+that corresponds the compiled program and the library facilities that
+it required.  You can execute this file directly using <tt class="docutils literal"><span class="pre">lli</span></tt> tool,
+compile it to native assembly with the <tt class="docutils literal"><span class="pre">llc</span></tt>, optimize or analyze it
+further with the <tt class="docutils literal"><span class="pre">opt</span></tt> tool, etc.</p>
+<p>Alternatively you can directly output an executable with clang with:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..<span class="p">></span> <span class="n">clang</span> hello.c -o hello.exe
+</pre></div>
+</div>
+<p>The <tt class="docutils literal"><span class="pre">-o</span> <span class="pre">hello.exe</span></tt> is required because clang currently outputs <tt class="docutils literal"><span class="pre">a.out</span></tt>
+when neither <tt class="docutils literal"><span class="pre">-o</span></tt> nor <tt class="docutils literal"><span class="pre">-c</span></tt> are given.</p>
+</li>
+<li><p class="first">Run the program using the just-in-time compiler:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..<span class="p">></span> <span class="n">lli</span> hello.bc
+</pre></div>
+</div>
+</li>
+<li><p class="first">Use the <tt class="docutils literal"><span class="pre">llvm-dis</span></tt> utility to take a look at the LLVM assembly code:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..<span class="p">></span> <span class="n">llvm</span>-dis <span class="p"><</span> <span class="n">hello</span>.bc <span class="p">|</span> <span class="n">more</span>
+</pre></div>
+</div>
+</li>
+<li><p class="first">Compile the program to object code using the LLC code generator:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..<span class="p">></span> <span class="n">llc</span> -filetype<span class="o">=</span>obj hello.bc
+</pre></div>
+</div>
+</li>
+<li><p class="first">Link to binary using Microsoft link:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..<span class="p">></span> <span class="n">link</span> hello.obj -defaultlib:libcmt
+</pre></div>
+</div>
+</li>
+<li><p class="first">Execute the native code program:</p>
+<div class="highlight-bat"><div class="highlight"><pre>C:\..<span class="p">></span> <span class="n">hello</span>.exe
+</pre></div>
+</div>
+</li>
+</ol>
+</div>
+<div class="section" id="common-problems">
+<h2><a class="toc-backref" href="#id8">Common Problems</a><a class="headerlink" href="#common-problems" title="Permalink to this headline">¶</a></h2>
+<p>If you are having problems building or using LLVM, or if you have any other
+general questions about LLVM, please consult the <a class="reference internal" href="FAQ.html"><em>Frequently Asked Questions</em></a> page.</p>
+</div>
+<div class="section" id="links">
+<h2><a class="toc-backref" href="#id9">Links</a><a class="headerlink" href="#links" title="Permalink to this headline">¶</a></h2>
+<p>This document is just an <strong>introduction</strong> to how to use LLVM to do some simple
+things... there are many more interesting and complicated things that you can
+do that aren’t documented here (but we’ll gladly accept a patch if you want to
+write something up!).  For more information about LLVM, check out:&