/usr/share/doc/llvm-3.7-doc/html/FAQ.html is in llvm-3.7-doc 1:3.7.1-2ubuntu2.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 | <!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 3.7 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: '3.7',
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 3.7 documentation" href="index.html" />
<link rel="next" title="The LLVM Lexicon" href="Lexicon.html" />
<link rel="prev" title="Building LLVM With Autotools" href="BuildingLLVMWithAutotools.html" />
<style type="text/css">
table.right { float: right; margin-left: 20px; }
table.right td { border: 1px solid #ccc; }
</style>
</head>
<body role="document">
<div class="logo">
<a href="index.html">
<img src="_static/logo.png"
alt="LLVM Logo" width="250" height="88"/></a>
</div>
<div class="related" role="navigation" aria-label="related navigation">
<h3>Navigation</h3>
<ul>
<li class="right" style="margin-right: 10px">
<a href="genindex.html" title="General Index"
accesskey="I">index</a></li>
<li class="right" >
<a href="Lexicon.html" title="The LLVM Lexicon"
accesskey="N">next</a> |</li>
<li class="right" >
<a href="BuildingLLVMWithAutotools.html" title="Building LLVM With Autotools"
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" role="main">
<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 <code class="docutils literal"><span class="pre">configure</span></code> 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 <code class="docutils literal"><span class="pre">GetElementPtr</span></code> 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 <code class="docutils literal"><span class="pre">llvm.global_ctors</span></code> and <code class="docutils literal"><span class="pre">_GLOBAL__I_a...</span></code> stuff that happens when I <code class="docutils literal"><span class="pre">#include</span> <span class="pre"><iostream></span></code>?</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 “<code class="docutils literal"><span class="pre">undef</span></code>” 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>, <code class="docutils literal"><span class="pre">%result</span></code> is just a name given to the <code class="docutils literal"><span class="pre">Value</span></code> of the <code class="docutils literal"><span class="pre">add</span></code>
instruction. In other words, <code class="docutils literal"><span class="pre">%result</span></code> <em>is</em> the add instruction. The
“assignment” doesn’t explicitly “store” anything to any “virtual register”;
the “<code class="docutils literal"><span class="pre">=</span></code>” 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 <code class="docutils literal"><span class="pre">Value</span></code>‘s that they
reference. In fact, the names of dummy numbered temporaries like <code class="docutils literal"><span class="pre">%1</span></code> are
not explicitly represented in the in-memory representation at all (see
<code class="docutils literal"><span class="pre">Value::getName()</span></code>).</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 <code class="docutils literal"><span class="pre">configure</span></code> script attempts to locate first <code class="docutils literal"><span class="pre">gcc</span></code> and then <code class="docutils literal"><span class="pre">cc</span></code>,
unless it finds compiler paths set in <code class="docutils literal"><span class="pre">CC</span></code> and <code class="docutils literal"><span class="pre">CXX</span></code> for the C and C++
compiler, respectively.</p>
<p>If <code class="docutils literal"><span class="pre">configure</span></code> finds the wrong compiler, either adjust your <code class="docutils literal"><span class="pre">PATH</span></code>
environment variable or set <code class="docutils literal"><span class="pre">CC</span></code> and <code class="docutils literal"><span class="pre">CXX</span></code> 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 <code class="docutils literal"><span class="pre">configure</span></code> 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 <code class="docutils literal"><span class="pre">configure</span></code> script uses the <code class="docutils literal"><span class="pre">PATH</span></code> 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 <code class="docutils literal"><span class="pre">PATH</span></code> environment variable so that the correct program
appears first in the <code class="docutils literal"><span class="pre">PATH</span></code>. This may work, but may not be convenient
when you want them <em>first</em> in your path for other work.</li>
<li>Run <code class="docutils literal"><span class="pre">configure</span></code> with an alternative <code class="docutils literal"><span class="pre">PATH</span></code> 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> $LLVM_SRC_DIR/configure ...
</pre></div>
</div>
<p>This is still somewhat inconvenient, but it allows <code class="docutils literal"><span class="pre">configure</span></code> to do its
work without having to adjust your <code class="docutils literal"><span class="pre">PATH</span></code> 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 <code class="docutils literal"><span class="pre">--disable-shared</span> <span class="pre">option</span></code>. 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><span class="p">;</span>
</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 <code class="docutils literal"><span class="pre">libtool</span></code>, <code class="docutils literal"><span class="pre">autoconf</span></code>, 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 <code class="docutils literal"><span class="pre">make</span> <span class="pre">clean</span></code> and then
<code class="docutils literal"><span class="pre">make</span></code> 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
<code class="docutils literal"><span class="pre">gmake</span></code> 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 <code class="docutils literal"><span class="pre">.d</span></code>
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> $LLVM_OBJ_DIR
<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 <code class="docutils literal"><span class="pre">make</span> <span class="pre">clean</span></code> 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 <code class="docutils literal"><span class="pre">GetElementPtr</span></code> 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 <code class="docutils literal"><span class="pre">sizeof</span></code>. It’s common for <code class="docutils literal"><span class="pre">sizeof(long)</span></code> to vary
between platforms. In most C front-ends, <code class="docutils literal"><span class="pre">sizeof</span></code> 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 <code class="docutils literal"><span class="pre">llvm.global_ctors</span></code> and <code class="docutils literal"><span class="pre">_GLOBAL__I_a...</span></code> stuff that happens when I <code class="docutils literal"><span class="pre">#include</span> <span class="pre"><iostream></span></code>?</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 <code class="docutils literal"><span class="pre">#include</span></code> the <code class="docutils literal"><span class="pre"><iostream></span></code> header into a C++ translation unit,
the file will probably use the <code class="docutils literal"><span class="pre">std::cin</span></code>/<code class="docutils literal"><span class="pre">std::cout</span></code>/... 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 <code class="docutils literal"><span class="pre">std::cout</span></code>, for example, the object would not necessarily be
automatically initialized before your use.</p>
<p>To make <code class="docutils literal"><span class="pre">std::cout</span></code> 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 <code class="docutils literal"><span class="pre"><iostream></span></code>. 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
<code class="docutils literal"><span class="pre">.ll</span></code> 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 <code class="docutils literal"><span class="pre">printf()</span></code> instead of
<code class="docutils literal"><span class="pre">iostream</span></code>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 <code class="docutils literal"><span class="pre">volatile</span></code> 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 “<code class="docutils literal"><span class="pre">undef</span></code>” 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><code class="docutils literal"><span class="pre">undef</span></code> 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 “<code class="docutils literal"><span class="pre">ret</span> <span class="pre">i32</span> <span class="pre">undef</span></code>” because “<code class="docutils literal"><span class="pre">i</span></code>” 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 “<code class="docutils literal"><span class="pre">opt</span> <span class="pre">-instcombine</span> <span class="pre">-simplifycfg</span></code>”. 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 <code class="docutils literal"><span class="pre">@foo</span></code>/<code class="docutils literal"><span class="pre">false</span></code> into <code class="docutils literal"><span class="pre">bar</span></code>, 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 <code class="docutils literal"><span class="pre">@foo</span></code>
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 <code class="docutils literal"><span class="pre">%X</span></code> was an input argument to <code class="docutils literal"><span class="pre">@test</span></code>, 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 <code class="docutils literal"><span class="pre">%X</span></code> <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
<code class="docutils literal"><span class="pre">instcombine</span></code>/<code class="docutils literal"><span class="pre">simplifycfg</span></code> 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 “<code class="docutils literal"><span class="pre">-inline</span> <span class="pre">-instcombine</span> <span class="pre">-simplifycfg</span></code>” 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" role="navigation" aria-label="related navigation">
<h3>Navigation</h3>
<ul>
<li class="right" style="margin-right: 10px">
<a href="genindex.html" title="General Index"
>index</a></li>
<li class="right" >
<a href="Lexicon.html" title="The LLVM Lexicon"
>next</a> |</li>
<li class="right" >
<a href="BuildingLLVMWithAutotools.html" title="Building LLVM With Autotools"
>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" role="contentinfo">
© Copyright 2003-2016, LLVM Project.
Last updated on 2016-04-18.
Created using <a href="http://sphinx-doc.org/">Sphinx</a> 1.3.6.
</div>
</body>
</html>
|