/usr/share/doc/llvm-4.0-doc/html/TestingGuide.html is in llvm-4.0-doc 1:4.0-1ubuntu1~16.04.2.
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 | <!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 Testing Infrastructure Guide — LLVM 4.0 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: '4.0',
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 4.0 documentation" href="index.html" />
<link rel="next" title="LLVM test-suite Guide" href="TestSuiteMakefileGuide.html" />
<link rel="prev" title="Code Reviews with Phabricator" href="Phabricator.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="TestSuiteMakefileGuide.html" title="LLVM test-suite Guide"
accesskey="N">next</a> |</li>
<li class="right" >
<a href="Phabricator.html" title="Code Reviews with Phabricator"
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="llvm-testing-infrastructure-guide">
<h1>LLVM Testing Infrastructure Guide<a class="headerlink" href="#llvm-testing-infrastructure-guide" 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="#requirements" id="id6">Requirements</a></li>
<li><a class="reference internal" href="#llvm-testing-infrastructure-organization" id="id7">LLVM testing infrastructure organization</a><ul>
<li><a class="reference internal" href="#regression-tests" id="id8">Regression tests</a></li>
<li><a class="reference internal" href="#test-suite" id="id9"><code class="docutils literal"><span class="pre">test-suite</span></code></a></li>
<li><a class="reference internal" href="#debugging-information-tests" id="id10">Debugging Information tests</a></li>
</ul>
</li>
<li><a class="reference internal" href="#quick-start" id="id11">Quick start</a><ul>
<li><a class="reference internal" href="#id1" id="id12">Regression tests</a></li>
<li><a class="reference internal" href="#id2" id="id13">Debugging Information tests</a></li>
</ul>
</li>
<li><a class="reference internal" href="#regression-test-structure" id="id14">Regression test structure</a><ul>
<li><a class="reference internal" href="#writing-new-regression-tests" id="id15">Writing new regression tests</a></li>
<li><a class="reference internal" href="#extra-files" id="id16">Extra files</a></li>
<li><a class="reference internal" href="#fragile-tests" id="id17">Fragile tests</a></li>
<li><a class="reference internal" href="#platform-specific-tests" id="id18">Platform-Specific Tests</a><ul>
<li><a class="reference internal" href="#requires-and-requires-any-directive" id="id19">REQUIRES and REQUIRES-ANY directive</a></li>
</ul>
</li>
<li><a class="reference internal" href="#substitutions" id="id20">Substitutions</a></li>
<li><a class="reference internal" href="#options" id="id21">Options</a></li>
<li><a class="reference internal" href="#other-features" id="id22">Other Features</a></li>
</ul>
</li>
<li><a class="reference internal" href="#test-suite-overview" id="id23"><code class="docutils literal"><span class="pre">test-suite</span></code> Overview</a><ul>
<li><a class="reference internal" href="#test-suite-quickstart" id="id24"><code class="docutils literal"><span class="pre">test-suite</span></code> Quickstart</a></li>
<li><a class="reference internal" href="#test-suite-makefiles" id="id25"><code class="docutils literal"><span class="pre">test-suite</span></code> Makefiles</a></li>
</ul>
</li>
</ul>
</div>
<div class="toctree-wrapper compound">
</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>This document is the reference manual for the LLVM testing
infrastructure. It documents the structure of the LLVM testing
infrastructure, the tools needed to use it, and how to add and run
tests.</p>
</div>
<div class="section" id="requirements">
<h2><a class="toc-backref" href="#id6">Requirements</a><a class="headerlink" href="#requirements" title="Permalink to this headline">¶</a></h2>
<p>In order to use the LLVM testing infrastructure, you will need all of the
software required to build LLVM, as well as <a class="reference external" href="http://python.org">Python</a> 2.7 or
later.</p>
<p>If you intend to run the <a class="reference internal" href="#test-suite-overview"><span>test-suite</span></a>, you will also
need a development version of zlib (zlib1g-dev is known to work on several Linux
distributions).</p>
</div>
<div class="section" id="llvm-testing-infrastructure-organization">
<h2><a class="toc-backref" href="#id7">LLVM testing infrastructure organization</a><a class="headerlink" href="#llvm-testing-infrastructure-organization" title="Permalink to this headline">¶</a></h2>
<p>The LLVM testing infrastructure contains two major categories of tests:
regression tests and whole programs. The regression tests are contained
inside the LLVM repository itself under <code class="docutils literal"><span class="pre">llvm/test</span></code> and are expected
to always pass – they should be run before every commit.</p>
<p>The whole programs tests are referred to as the “LLVM test suite” (or
“test-suite”) and are in the <code class="docutils literal"><span class="pre">test-suite</span></code> module in subversion. For
historical reasons, these tests are also referred to as the “nightly
tests” in places, which is less ambiguous than “test-suite” and remains
in use although we run them much more often than nightly.</p>
<div class="section" id="regression-tests">
<h3><a class="toc-backref" href="#id8">Regression tests</a><a class="headerlink" href="#regression-tests" title="Permalink to this headline">¶</a></h3>
<p>The regression tests are small pieces of code that test a specific
feature of LLVM or trigger a specific bug in LLVM. The language they are
written in depends on the part of LLVM being tested. These tests are driven by
the <a class="reference internal" href="CommandGuide/lit.html"><em>Lit</em></a> testing tool (which is part of LLVM), and
are located in the <code class="docutils literal"><span class="pre">llvm/test</span></code> directory.</p>
<p>Typically when a bug is found in LLVM, a regression test containing just
enough code to reproduce the problem should be written and placed
somewhere underneath this directory. For example, it can be a small
piece of LLVM IR distilled from an actual application or benchmark.</p>
</div>
<div class="section" id="test-suite">
<h3><a class="toc-backref" href="#id9"><code class="docutils literal"><span class="pre">test-suite</span></code></a><a class="headerlink" href="#test-suite" title="Permalink to this headline">¶</a></h3>
<p>The test suite contains whole programs, which are pieces of code which
can be compiled and linked into a stand-alone program that can be
executed. These programs are generally written in high level languages
such as C or C++.</p>
<p>These programs are compiled using a user specified compiler and set of
flags, and then executed to capture the program output and timing
information. The output of these programs is compared to a reference
output to ensure that the program is being compiled correctly.</p>
<p>In addition to compiling and executing programs, whole program tests
serve as a way of benchmarking LLVM performance, both in terms of the
efficiency of the programs generated as well as the speed with which
LLVM compiles, optimizes, and generates code.</p>
<p>The test-suite is located in the <code class="docutils literal"><span class="pre">test-suite</span></code> Subversion module.</p>
</div>
<div class="section" id="debugging-information-tests">
<h3><a class="toc-backref" href="#id10">Debugging Information tests</a><a class="headerlink" href="#debugging-information-tests" title="Permalink to this headline">¶</a></h3>
<p>The test suite contains tests to check quality of debugging information.
The test are written in C based languages or in LLVM assembly language.</p>
<p>These tests are compiled and run under a debugger. The debugger output
is checked to validate of debugging information. See README.txt in the
test suite for more information . This test suite is located in the
<code class="docutils literal"><span class="pre">debuginfo-tests</span></code> Subversion module.</p>
</div>
</div>
<div class="section" id="quick-start">
<h2><a class="toc-backref" href="#id11">Quick start</a><a class="headerlink" href="#quick-start" title="Permalink to this headline">¶</a></h2>
<p>The tests are located in two separate Subversion modules. The
regressions tests are in the main “llvm” module under the directory
<code class="docutils literal"><span class="pre">llvm/test</span></code> (so you get these tests for free with the main LLVM tree).
Use <code class="docutils literal"><span class="pre">make</span> <span class="pre">check-all</span></code> to run the regression tests after building LLVM.</p>
<p>The more comprehensive test suite that includes whole programs in C and C++
is in the <code class="docutils literal"><span class="pre">test-suite</span></code> module. See <a class="reference internal" href="#test-suite-quickstart"><span>test-suite Quickstart</span></a> for more information on running these tests.</p>
<div class="section" id="id1">
<h3><a class="toc-backref" href="#id12">Regression tests</a><a class="headerlink" href="#id1" title="Permalink to this headline">¶</a></h3>
<p>To run all of the LLVM regression tests use the check-llvm target:</p>
<div class="highlight-bash"><div class="highlight"><pre>% make check-llvm
</pre></div>
</div>
<p>If you have <a class="reference external" href="http://clang.llvm.org/">Clang</a> checked out and built, you
can run the LLVM and Clang tests simultaneously using:</p>
<div class="highlight-bash"><div class="highlight"><pre>% make check-all
</pre></div>
</div>
<p>To run the tests with Valgrind (Memcheck by default), use the <code class="docutils literal"><span class="pre">LIT_ARGS</span></code> make
variable to pass the required options to lit. For example, you can use:</p>
<div class="highlight-bash"><div class="highlight"><pre>% make check <span class="nv">LIT_ARGS</span><span class="o">=</span><span class="s2">"-v --vg --vg-leak"</span>
</pre></div>
</div>
<p>to enable testing with valgrind and with leak checking enabled.</p>
<p>To run individual tests or subsets of tests, you can use the <code class="docutils literal"><span class="pre">llvm-lit</span></code>
script which is built as part of LLVM. For example, to run the
<code class="docutils literal"><span class="pre">Integer/BitPacked.ll</span></code> test by itself you can run:</p>
<div class="highlight-bash"><div class="highlight"><pre>% llvm-lit ~/llvm/test/Integer/BitPacked.ll
</pre></div>
</div>
<p>or to run all of the ARM CodeGen tests:</p>
<div class="highlight-bash"><div class="highlight"><pre>% llvm-lit ~/llvm/test/CodeGen/ARM
</pre></div>
</div>
<p>For more information on using the <strong class="program">lit</strong> tool, see <code class="docutils literal"><span class="pre">llvm-lit</span> <span class="pre">--help</span></code>
or the <a class="reference internal" href="CommandGuide/lit.html"><em>lit man page</em></a>.</p>
</div>
<div class="section" id="id2">
<h3><a class="toc-backref" href="#id13">Debugging Information tests</a><a class="headerlink" href="#id2" title="Permalink to this headline">¶</a></h3>
<p>To run debugging information tests simply checkout the tests inside
clang/test directory.</p>
<div class="highlight-bash"><div class="highlight"><pre>% <span class="nb">cd</span> clang/test
% svn co http://llvm.org/svn/llvm-project/debuginfo-tests/trunk debuginfo-tests
</pre></div>
</div>
<p>These tests are already set up to run as part of clang regression tests.</p>
</div>
</div>
<div class="section" id="regression-test-structure">
<h2><a class="toc-backref" href="#id14">Regression test structure</a><a class="headerlink" href="#regression-test-structure" title="Permalink to this headline">¶</a></h2>
<p>The LLVM regression tests are driven by <strong class="program">lit</strong> and are located in the
<code class="docutils literal"><span class="pre">llvm/test</span></code> directory.</p>
<p>This directory contains a large array of small tests that exercise
various features of LLVM and to ensure that regressions do not occur.
The directory is broken into several sub-directories, each focused on a
particular area of LLVM.</p>
<div class="section" id="writing-new-regression-tests">
<h3><a class="toc-backref" href="#id15">Writing new regression tests</a><a class="headerlink" href="#writing-new-regression-tests" title="Permalink to this headline">¶</a></h3>
<p>The regression test structure is very simple, but does require some
information to be set. This information is gathered via <code class="docutils literal"><span class="pre">configure</span></code>
and is written to a file, <code class="docutils literal"><span class="pre">test/lit.site.cfg</span></code> in the build directory.
The <code class="docutils literal"><span class="pre">llvm/test</span></code> Makefile does this work for you.</p>
<p>In order for the regression tests to work, each directory of tests must
have a <code class="docutils literal"><span class="pre">lit.local.cfg</span></code> file. <strong class="program">lit</strong> looks for this file to determine
how to run the tests. This file is just Python code and thus is very
flexible, but we’ve standardized it for the LLVM regression tests. If
you’re adding a directory of tests, just copy <code class="docutils literal"><span class="pre">lit.local.cfg</span></code> from
another directory to get running. The standard <code class="docutils literal"><span class="pre">lit.local.cfg</span></code> simply
specifies which files to look in for tests. Any directory that contains
only directories does not need the <code class="docutils literal"><span class="pre">lit.local.cfg</span></code> file. Read the <a class="reference internal" href="CommandGuide/lit.html"><em>Lit
documentation</em></a> for more information.</p>
<p>Each test file must contain lines starting with “RUN:” that tell <strong class="program">lit</strong>
how to run it. If there are no RUN lines, <strong class="program">lit</strong> will issue an error
while running a test.</p>
<p>RUN lines are specified in the comments of the test program using the
keyword <code class="docutils literal"><span class="pre">RUN</span></code> followed by a colon, and lastly the command (pipeline)
to execute. Together, these lines form the “script” that <strong class="program">lit</strong>
executes to run the test case. The syntax of the RUN lines is similar to a
shell’s syntax for pipelines including I/O redirection and variable
substitution. However, even though these lines may <em>look</em> like a shell
script, they are not. RUN lines are interpreted by <strong class="program">lit</strong>.
Consequently, the syntax differs from shell in a few ways. You can specify
as many RUN lines as needed.</p>
<p><strong class="program">lit</strong> performs substitution on each RUN line to replace LLVM tool names
with the full paths to the executable built for each tool (in
<code class="docutils literal"><span class="pre">$(LLVM_OBJ_ROOT)/$(BuildMode)/bin)</span></code>. This ensures that <strong class="program">lit</strong> does
not invoke any stray LLVM tools in the user’s path during testing.</p>
<p>Each RUN line is executed on its own, distinct from other lines unless
its last character is <code class="docutils literal"><span class="pre">\</span></code>. This continuation character causes the RUN
line to be concatenated with the next one. In this way you can build up
long pipelines of commands without making huge line lengths. The lines
ending in <code class="docutils literal"><span class="pre">\</span></code> are concatenated until a RUN line that doesn’t end in
<code class="docutils literal"><span class="pre">\</span></code> is found. This concatenated set of RUN lines then constitutes one
execution. <strong class="program">lit</strong> will substitute variables and arrange for the pipeline
to be executed. If any process in the pipeline fails, the entire line (and
test case) fails too.</p>
<p>Below is an example of legal RUN lines in a <code class="docutils literal"><span class="pre">.ll</span></code> file:</p>
<div class="highlight-llvm"><div class="highlight"><pre><span class="c">; RUN: llvm-as < %s | llvm-dis > %t1</span>
<span class="c">; RUN: llvm-dis < %s.bc-13 > %t2</span>
<span class="c">; RUN: diff %t1 %t2</span>
</pre></div>
</div>
<p>As with a Unix shell, the RUN lines permit pipelines and I/O
redirection to be used.</p>
<p>There are some quoting rules that you must pay attention to when writing
your RUN lines. In general nothing needs to be quoted. <strong class="program">lit</strong> won’t
strip off any quote characters so they will get passed to the invoked program.
To avoid this use curly braces to tell <strong class="program">lit</strong> that it should treat
everything enclosed as one value.</p>
<p>In general, you should strive to keep your RUN lines as simple as possible,
using them only to run tools that generate textual output you can then examine.
The recommended way to examine output to figure out if the test passes is using
the <a class="reference internal" href="CommandGuide/FileCheck.html"><em>FileCheck tool</em></a>. <em>[The usage of grep in RUN
lines is deprecated - please do not send or commit patches that use it.]</em></p>
<p>Put related tests into a single file rather than having a separate file per
test. Check if there are files already covering your feature and consider
adding your code there instead of creating a new file.</p>
</div>
<div class="section" id="extra-files">
<h3><a class="toc-backref" href="#id16">Extra files</a><a class="headerlink" href="#extra-files" title="Permalink to this headline">¶</a></h3>
<p>If your test requires extra files besides the file containing the <code class="docutils literal"><span class="pre">RUN:</span></code>
lines, the idiomatic place to put them is in a subdirectory <code class="docutils literal"><span class="pre">Inputs</span></code>.
You can then refer to the extra files as <code class="docutils literal"><span class="pre">%S/Inputs/foo.bar</span></code>.</p>
<p>For example, consider <code class="docutils literal"><span class="pre">test/Linker/ident.ll</span></code>. The directory structure is
as follows:</p>
<div class="highlight-python"><div class="highlight"><pre>test/
Linker/
ident.ll
Inputs/
ident.a.ll
ident.b.ll
</pre></div>
</div>
<p>For convenience, these are the contents:</p>
<div class="highlight-llvm"><div class="highlight"><pre><span class="c">;;;;; ident.ll:</span>
<span class="c">; RUN: llvm-link %S/Inputs/ident.a.ll %S/Inputs/ident.b.ll -S | FileCheck %s</span>
<span class="c">; Verify that multiple input llvm.ident metadata are linked together.</span>
<span class="c">; CHECK-DAG: !llvm.ident = !{!0, !1, !2}</span>
<span class="c">; CHECK-DAG: "Compiler V1"</span>
<span class="c">; CHECK-DAG: "Compiler V2"</span>
<span class="c">; CHECK-DAG: "Compiler V3"</span>
<span class="c">;;;;; Inputs/ident.a.ll:</span>
<span class="nv">!llvm.ident</span> <span class="p">=</span> <span class="p">!{</span><span class="nv nv-Anonymous">!0</span><span class="p">,</span> <span class="nv nv-Anonymous">!1</span><span class="p">}</span>
<span class="nv nv-Anonymous">!0</span> <span class="p">=</span> <span class="kt">metadata</span> <span class="p">!{</span><span class="kt">metadata</span> <span class="nv">!"Compiler V1"</span><span class="p">}</span>
<span class="nv nv-Anonymous">!1</span> <span class="p">=</span> <span class="kt">metadata</span> <span class="p">!{</span><span class="kt">metadata</span> <span class="nv">!"Compiler V2"</span><span class="p">}</span>
<span class="c">;;;;; Inputs/ident.b.ll:</span>
<span class="nv">!llvm.ident</span> <span class="p">=</span> <span class="p">!{</span><span class="nv nv-Anonymous">!0</span><span class="p">}</span>
<span class="nv nv-Anonymous">!0</span> <span class="p">=</span> <span class="kt">metadata</span> <span class="p">!{</span><span class="kt">metadata</span> <span class="nv">!"Compiler V3"</span><span class="p">}</span>
</pre></div>
</div>
<p>For symmetry reasons, <code class="docutils literal"><span class="pre">ident.ll</span></code> is just a dummy file that doesn’t
actually participate in the test besides holding the <code class="docutils literal"><span class="pre">RUN:</span></code> lines.</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">Some existing tests use <code class="docutils literal"><span class="pre">RUN:</span> <span class="pre">true</span></code> in extra files instead of just
putting the extra files in an <code class="docutils literal"><span class="pre">Inputs/</span></code> directory. This pattern is
deprecated.</p>
</div>
</div>
<div class="section" id="fragile-tests">
<h3><a class="toc-backref" href="#id17">Fragile tests</a><a class="headerlink" href="#fragile-tests" title="Permalink to this headline">¶</a></h3>
<p>It is easy to write a fragile test that would fail spuriously if the tool being
tested outputs a full path to the input file. For example, <strong class="program">opt</strong> by
default outputs a <code class="docutils literal"><span class="pre">ModuleID</span></code>:</p>
<div class="highlight-console"><div class="highlight"><pre><span class="gp">$</span> cat example.ll
<span class="go">define i32 @main() nounwind {</span>
<span class="go"> ret i32 0</span>
<span class="go">}</span>
<span class="gp">$</span> opt -S /path/to/example.ll
<span class="go">; ModuleID = '/path/to/example.ll'</span>
<span class="go">define i32 @main() nounwind {</span>
<span class="go"> ret i32 0</span>
<span class="go">}</span>
</pre></div>
</div>
<p><code class="docutils literal"><span class="pre">ModuleID</span></code> can unexpetedly match against <code class="docutils literal"><span class="pre">CHECK</span></code> lines. For example:</p>
<div class="highlight-llvm"><div class="highlight"><pre><span class="c">; RUN: opt -S %s | FileCheck</span>
<span class="k">define</span> <span class="k">i32</span> <span class="vg">@main</span><span class="p">()</span> <span class="k">nounwind</span> <span class="p">{</span>
<span class="c">; CHECK-NOT: load</span>
<span class="k">ret</span> <span class="k">i32</span> <span class="m">0</span>
<span class="p">}</span>
</pre></div>
</div>
<p>This test will fail if placed into a <code class="docutils literal"><span class="pre">download</span></code> directory.</p>
<p>To make your tests robust, always use <code class="docutils literal"><span class="pre">opt</span> <span class="pre">...</span> <span class="pre"><</span> <span class="pre">%s</span></code> in the RUN line.
<strong class="program">opt</strong> does not output a <code class="docutils literal"><span class="pre">ModuleID</span></code> when input comes from stdin.</p>
</div>
<div class="section" id="platform-specific-tests">
<h3><a class="toc-backref" href="#id18">Platform-Specific Tests</a><a class="headerlink" href="#platform-specific-tests" title="Permalink to this headline">¶</a></h3>
<p>Whenever adding tests that require the knowledge of a specific platform,
either related to code generated, specific output or back-end features,
you must make sure to isolate the features, so that buildbots that
run on different architectures (and don’t even compile all back-ends),
don’t fail.</p>
<p>The first problem is to check for target-specific output, for example sizes
of structures, paths and architecture names, for example:</p>
<ul class="simple">
<li>Tests containing Windows paths will fail on Linux and vice-versa.</li>
<li>Tests that check for <code class="docutils literal"><span class="pre">x86_64</span></code> somewhere in the text will fail anywhere else.</li>
<li>Tests where the debug information calculates the size of types and structures.</li>
</ul>
<p>Also, if the test rely on any behaviour that is coded in any back-end, it must
go in its own directory. So, for instance, code generator tests for ARM go
into <code class="docutils literal"><span class="pre">test/CodeGen/ARM</span></code> and so on. Those directories contain a special
<code class="docutils literal"><span class="pre">lit</span></code> configuration file that ensure all tests in that directory will
only run if a specific back-end is compiled and available.</p>
<p>For instance, on <code class="docutils literal"><span class="pre">test/CodeGen/ARM</span></code>, the <code class="docutils literal"><span class="pre">lit.local.cfg</span></code> is:</p>
<div class="highlight-python"><div class="highlight"><pre><span class="n">config</span><span class="o">.</span><span class="n">suffixes</span> <span class="o">=</span> <span class="p">[</span><span class="s1">'.ll'</span><span class="p">,</span> <span class="s1">'.c'</span><span class="p">,</span> <span class="s1">'.cpp'</span><span class="p">,</span> <span class="s1">'.test'</span><span class="p">]</span>
<span class="k">if</span> <span class="ow">not</span> <span class="s1">'ARM'</span> <span class="ow">in</span> <span class="n">config</span><span class="o">.</span><span class="n">root</span><span class="o">.</span><span class="n">targets</span><span class="p">:</span>
<span class="n">config</span><span class="o">.</span><span class="n">unsupported</span> <span class="o">=</span> <span class="bp">True</span>
</pre></div>
</div>
<p>Other platform-specific tests are those that depend on a specific feature
of a specific sub-architecture, for example only to Intel chips that support <code class="docutils literal"><span class="pre">AVX2</span></code>.</p>
<p>For instance, <code class="docutils literal"><span class="pre">test/CodeGen/X86/psubus.ll</span></code> tests three sub-architecture
variants:</p>
<div class="highlight-llvm"><div class="highlight"><pre><span class="c">; RUN: llc -mcpu=core2 < %s | FileCheck %s -check-prefix=SSE2</span>
<span class="c">; RUN: llc -mcpu=corei7-avx < %s | FileCheck %s -check-prefix=AVX1</span>
<span class="c">; RUN: llc -mcpu=core-avx2 < %s | FileCheck %s -check-prefix=AVX2</span>
</pre></div>
</div>
<p>And the checks are different:</p>
<div class="highlight-llvm"><div class="highlight"><pre><span class="c">; SSE2: @test1</span>
<span class="c">; SSE2: psubusw LCPI0_0(%rip), %xmm0</span>
<span class="c">; AVX1: @test1</span>
<span class="c">; AVX1: vpsubusw LCPI0_0(%rip), %xmm0, %xmm0</span>
<span class="c">; AVX2: @test1</span>
<span class="c">; AVX2: vpsubusw LCPI0_0(%rip), %xmm0, %xmm0</span>
</pre></div>
</div>
<p>So, if you’re testing for a behaviour that you know is platform-specific or
depends on special features of sub-architectures, you must add the specific
triple, test with the specific FileCheck and put it into the specific
directory that will filter out all other architectures.</p>
<div class="section" id="requires-and-requires-any-directive">
<h4><a class="toc-backref" href="#id19">REQUIRES and REQUIRES-ANY directive</a><a class="headerlink" href="#requires-and-requires-any-directive" title="Permalink to this headline">¶</a></h4>
<p>Some tests can be enabled only in specific situation - like having
debug build. Use <code class="docutils literal"><span class="pre">REQUIRES</span></code> directive to specify those requirements.</p>
<div class="highlight-llvm"><div class="highlight"><pre><span class="c">; This test will be only enabled in the build with asserts</span>
<span class="c">; REQUIRES: asserts</span>
</pre></div>
</div>
<p>You can separate requirements by a comma.
<code class="docutils literal"><span class="pre">REQUIRES</span></code> means all listed requirements must be satisfied.
<code class="docutils literal"><span class="pre">REQUIRES-ANY</span></code> means at least one must be satisfied.</p>
<p>List of features that can be used in <code class="docutils literal"><span class="pre">REQUIRES</span></code> and <code class="docutils literal"><span class="pre">REQUIRES-ANY</span></code> can be
found in lit.cfg files.</p>
</div>
</div>
<div class="section" id="substitutions">
<h3><a class="toc-backref" href="#id20">Substitutions</a><a class="headerlink" href="#substitutions" title="Permalink to this headline">¶</a></h3>
<p>Besides replacing LLVM tool names the following substitutions are performed in
RUN lines:</p>
<dl class="docutils">
<dt><code class="docutils literal"><span class="pre">%%</span></code></dt>
<dd>Replaced by a single <code class="docutils literal"><span class="pre">%</span></code>. This allows escaping other substitutions.</dd>
<dt><code class="docutils literal"><span class="pre">%s</span></code></dt>
<dd><p class="first">File path to the test case’s source. This is suitable for passing on the
command line as the input to an LLVM tool.</p>
<p class="last">Example: <code class="docutils literal"><span class="pre">/home/user/llvm/test/MC/ELF/foo_test.s</span></code></p>
</dd>
<dt><code class="docutils literal"><span class="pre">%S</span></code></dt>
<dd><p class="first">Directory path to the test case’s source.</p>
<p class="last">Example: <code class="docutils literal"><span class="pre">/home/user/llvm/test/MC/ELF</span></code></p>
</dd>
<dt><code class="docutils literal"><span class="pre">%t</span></code></dt>
<dd><p class="first">File path to a temporary file name that could be used for this test case.
The file name won’t conflict with other test cases. You can append to it
if you need multiple temporaries. This is useful as the destination of
some redirected output.</p>
<p class="last">Example: <code class="docutils literal"><span class="pre">/home/user/llvm.build/test/MC/ELF/Output/foo_test.s.tmp</span></code></p>
</dd>
<dt><code class="docutils literal"><span class="pre">%T</span></code></dt>
<dd><p class="first">Directory of <code class="docutils literal"><span class="pre">%t</span></code>.</p>
<p class="last">Example: <code class="docutils literal"><span class="pre">/home/user/llvm.build/test/MC/ELF/Output</span></code></p>
</dd>
</dl>
<p><code class="docutils literal"><span class="pre">%{pathsep}</span></code></p>
<blockquote>
<div>Expands to the path separator, i.e. <code class="docutils literal"><span class="pre">:</span></code> (or <code class="docutils literal"><span class="pre">;</span></code> on Windows).</div></blockquote>
<p><strong>LLVM-specific substitutions:</strong></p>
<dl class="docutils">
<dt><code class="docutils literal"><span class="pre">%shlibext</span></code></dt>
<dd><p class="first">The suffix for the host platforms shared library files. This includes the
period as the first character.</p>
<p class="last">Example: <code class="docutils literal"><span class="pre">.so</span></code> (Linux), <code class="docutils literal"><span class="pre">.dylib</span></code> (OS X), <code class="docutils literal"><span class="pre">.dll</span></code> (Windows)</p>
</dd>
<dt><code class="docutils literal"><span class="pre">%exeext</span></code></dt>
<dd><p class="first">The suffix for the host platforms executable files. This includes the
period as the first character.</p>
<p class="last">Example: <code class="docutils literal"><span class="pre">.exe</span></code> (Windows), empty on Linux.</p>
</dd>
<dt><code class="docutils literal"><span class="pre">%(line)</span></code>, <code class="docutils literal"><span class="pre">%(line+<number>)</span></code>, <code class="docutils literal"><span class="pre">%(line-<number>)</span></code></dt>
<dd>The number of the line where this substitution is used, with an optional
integer offset. This can be used in tests with multiple RUN lines, which
reference test file’s line numbers.</dd>
</dl>
<p><strong>Clang-specific substitutions:</strong></p>
<dl class="docutils">
<dt><code class="docutils literal"><span class="pre">%clang</span></code></dt>
<dd>Invokes the Clang driver.</dd>
<dt><code class="docutils literal"><span class="pre">%clang_cpp</span></code></dt>
<dd>Invokes the Clang driver for C++.</dd>
<dt><code class="docutils literal"><span class="pre">%clang_cl</span></code></dt>
<dd>Invokes the CL-compatible Clang driver.</dd>
<dt><code class="docutils literal"><span class="pre">%clangxx</span></code></dt>
<dd>Invokes the G++-compatible Clang driver.</dd>
<dt><code class="docutils literal"><span class="pre">%clang_cc1</span></code></dt>
<dd>Invokes the Clang frontend.</dd>
<dt><code class="docutils literal"><span class="pre">%itanium_abi_triple</span></code>, <code class="docutils literal"><span class="pre">%ms_abi_triple</span></code></dt>
<dd>These substitutions can be used to get the current target triple adjusted to
the desired ABI. For example, if the test suite is running with the
<code class="docutils literal"><span class="pre">i686-pc-win32</span></code> target, <code class="docutils literal"><span class="pre">%itanium_abi_triple</span></code> will expand to
<code class="docutils literal"><span class="pre">i686-pc-mingw32</span></code>. This allows a test to run with a specific ABI without
constraining it to a specific triple.</dd>
</dl>
<p>To add more substituations, look at <code class="docutils literal"><span class="pre">test/lit.cfg</span></code> or <code class="docutils literal"><span class="pre">lit.local.cfg</span></code>.</p>
</div>
<div class="section" id="options">
<h3><a class="toc-backref" href="#id21">Options</a><a class="headerlink" href="#options" title="Permalink to this headline">¶</a></h3>
<p>The llvm lit configuration allows to customize some things with user options:</p>
<dl class="docutils">
<dt><code class="docutils literal"><span class="pre">llc</span></code>, <code class="docutils literal"><span class="pre">opt</span></code>, ...</dt>
<dd><p class="first">Substitute the respective llvm tool name with a custom command line. This
allows to specify custom paths and default arguments for these tools.
Example:</p>
<p class="last">% llvm-lit “-Dllc=llc -verify-machineinstrs”</p>
</dd>
<dt><code class="docutils literal"><span class="pre">run_long_tests</span></code></dt>
<dd>Enable the execution of long running tests.</dd>
<dt><code class="docutils literal"><span class="pre">llvm_site_config</span></code></dt>
<dd>Load the specified lit configuration instead of the default one.</dd>
</dl>
</div>
<div class="section" id="other-features">
<h3><a class="toc-backref" href="#id22">Other Features</a><a class="headerlink" href="#other-features" title="Permalink to this headline">¶</a></h3>
<p>To make RUN line writing easier, there are several helper programs. These
helpers are in the PATH when running tests, so you can just call them using
their name. For example:</p>
<dl class="docutils">
<dt><code class="docutils literal"><span class="pre">not</span></code></dt>
<dd>This program runs its arguments and then inverts the result code from it.
Zero result codes become 1. Non-zero result codes become 0.</dd>
</dl>
<p>Sometimes it is necessary to mark a test case as “expected fail” or
XFAIL. You can easily mark a test as XFAIL just by including <code class="docutils literal"><span class="pre">XFAIL:</span></code>
on a line near the top of the file. This signals that the test case
should succeed if the test fails. Such test cases are counted separately
by the testing tool. To specify an expected fail, use the XFAIL keyword
in the comments of the test program followed by a colon and one or more
failure patterns. Each failure pattern can be either <code class="docutils literal"><span class="pre">*</span></code> (to specify
fail everywhere), or a part of a target triple (indicating the test
should fail on that platform), or the name of a configurable feature
(for example, <code class="docutils literal"><span class="pre">loadable_module</span></code>). If there is a match, the test is
expected to fail. If not, the test is expected to succeed. To XFAIL
everywhere just specify <code class="docutils literal"><span class="pre">XFAIL:</span> <span class="pre">*</span></code>. Here is an example of an <code class="docutils literal"><span class="pre">XFAIL</span></code>
line:</p>
<div class="highlight-llvm"><div class="highlight"><pre><span class="c">; XFAIL: darwin,sun</span>
</pre></div>
</div>
<p>To make the output more useful, <strong class="program">lit</strong> will scan
the lines of the test case for ones that contain a pattern that matches
<code class="docutils literal"><span class="pre">PR[0-9]+</span></code>. This is the syntax for specifying a PR (Problem Report) number
that is related to the test case. The number after “PR” specifies the
LLVM bugzilla number. When a PR number is specified, it will be used in
the pass/fail reporting. This is useful to quickly get some context when
a test fails.</p>
<p>Finally, any line that contains “END.” will cause the special
interpretation of lines to terminate. This is generally done right after
the last RUN: line. This has two side effects:</p>
<ol class="loweralpha simple">
<li>it prevents special interpretation of lines that are part of the test
program, not the instructions to the test case, and</li>
<li>it speeds things up for really big test cases by avoiding
interpretation of the remainder of the file.</li>
</ol>
</div>
</div>
<div class="section" id="test-suite-overview">
<span id="id3"></span><h2><a class="toc-backref" href="#id23"><code class="docutils literal"><span class="pre">test-suite</span></code> Overview</a><a class="headerlink" href="#test-suite-overview" title="Permalink to this headline">¶</a></h2>
<p>The <code class="docutils literal"><span class="pre">test-suite</span></code> module contains a number of programs that can be
compiled and executed. The <code class="docutils literal"><span class="pre">test-suite</span></code> includes reference outputs for
all of the programs, so that the output of the executed program can be
checked for correctness.</p>
<p><code class="docutils literal"><span class="pre">test-suite</span></code> tests are divided into three types of tests: MultiSource,
SingleSource, and External.</p>
<ul>
<li><p class="first"><code class="docutils literal"><span class="pre">test-suite/SingleSource</span></code></p>
<p>The SingleSource directory contains test programs that are only a
single source file in size. These are usually small benchmark
programs or small programs that calculate a particular value. Several
such programs are grouped together in each directory.</p>
</li>
<li><p class="first"><code class="docutils literal"><span class="pre">test-suite/MultiSource</span></code></p>
<p>The MultiSource directory contains subdirectories which contain
entire programs with multiple source files. Large benchmarks and
whole applications go here.</p>
</li>
<li><p class="first"><code class="docutils literal"><span class="pre">test-suite/External</span></code></p>
<p>The External directory contains Makefiles for building code that is
external to (i.e., not distributed with) LLVM. The most prominent
members of this directory are the SPEC 95 and SPEC 2000 benchmark
suites. The <code class="docutils literal"><span class="pre">External</span></code> directory does not contain these actual
tests, but only the Makefiles that know how to properly compile these
programs from somewhere else. When using <code class="docutils literal"><span class="pre">LNT</span></code>, use the
<code class="docutils literal"><span class="pre">--test-externals</span></code> option to include these tests in the results.</p>
</li>
</ul>
<div class="section" id="test-suite-quickstart">
<span id="id4"></span><h3><a class="toc-backref" href="#id24"><code class="docutils literal"><span class="pre">test-suite</span></code> Quickstart</a><a class="headerlink" href="#test-suite-quickstart" title="Permalink to this headline">¶</a></h3>
<p>The modern way of running the <code class="docutils literal"><span class="pre">test-suite</span></code> is focused on testing and
benchmarking complete compilers using the
<a class="reference external" href="http://llvm.org/docs/lnt">LNT</a> testing infrastructure.</p>
<p>For more information on using LNT to execute the <code class="docutils literal"><span class="pre">test-suite</span></code>, please
see the <a class="reference external" href="http://llvm.org/docs/lnt/quickstart.html">LNT Quickstart</a>
documentation.</p>
</div>
<div class="section" id="test-suite-makefiles">
<h3><a class="toc-backref" href="#id25"><code class="docutils literal"><span class="pre">test-suite</span></code> Makefiles</a><a class="headerlink" href="#test-suite-makefiles" title="Permalink to this headline">¶</a></h3>
<p>Historically, the <code class="docutils literal"><span class="pre">test-suite</span></code> was executed using a complicated setup
of Makefiles. The LNT based approach above is recommended for most
users, but there are some testing scenarios which are not supported by
the LNT approach. In addition, LNT currently uses the Makefile setup
under the covers and so developers who are interested in how LNT works
under the hood may want to understand the Makefile based setup.</p>
<p>For more information on the <code class="docutils literal"><span class="pre">test-suite</span></code> Makefile setup, please see
the <a class="reference internal" href="TestSuiteMakefileGuide.html"><em>Test Suite Makefile Guide</em></a>.</p>
</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="TestSuiteMakefileGuide.html" title="LLVM test-suite Guide"
>next</a> |</li>
<li class="right" >
<a href="Phabricator.html" title="Code Reviews with Phabricator"
>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 2003-2017, LLVM Project.
Last updated on 2017-08-02.
Created using <a href="http://sphinx-doc.org/">Sphinx</a> 1.3.6.
</div>
</body>
</html>
|