Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / f17a25c88b892d30c2b41ba7ecdfbdfb2b4be9cc / . / docs / TestingGuide.html
blob: cb880373de326ebea7566a4c56375f9d01e98a25 [file] [log] [blame]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame^]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
5 <title>LLVM Test Suite Guide</title>
6 <link rel="stylesheet" href="llvm.css" type="text/css">
7</head>
8<body>
9
10<div class="doc_title">
11 LLVM Test Suite Guide
12</div>
13
14<ol>
15 <li><a href="#overview">Overview</a></li>
16 <li><a href="#Requirements">Requirements</a></li>
17 <li><a href="#quick">Quick Start</a></li>
18 <li><a href="#org">LLVM Test Suite Organization</a>
19 <ul>
20 <li><a href="#codefragments">Code Fragments</a></li>
21 <li><a href="#wholeprograms">Whole Programs</a></li>
22 </ul>
23 </li>
24 <li><a href="#tree">LLVM Test Suite Tree</a></li>
25 <li><a href="#dgstructure">DejaGNU Structure</a></li>
26 <li><a href="#progstructure"><tt>llvm-test</tt> Structure</a></li>
27 <li><a href="#run">Running the LLVM Tests</a>
28 <ul>
29 <li><a href="#customtest">Writing custom tests for llvm-test</a></li>
30 </ul>
31 </li>
32 <li><a href="#nightly">Running the nightly tester</a></li>
33</ol>
34
35<div class="doc_author">
36 <p>Written by John T. Criswell, <a
37 href="http://llvm.x10sys.com/rspencer">Reid Spencer</a>, and Tanya Lattner</p>
38</div>
39
40<!--=========================================================================-->
41<div class="doc_section"><a name="overview">Overview</a></div>
42<!--=========================================================================-->
43
44<div class="doc_text">
45
46<p>This document is the reference manual for the LLVM test suite. It documents
47the structure of the LLVM test suite, the tools needed to use it, and how to add
48and run tests.</p>
49
50</div>
51
52<!--=========================================================================-->
53<div class="doc_section"><a name="Requirements">Requirements</a></div>
54<!--=========================================================================-->
55
56<div class="doc_text">
57
58<p>In order to use the LLVM test suite, you will need all of the software
59required to build LLVM, plus the following:</p>
60
61<dl>
62<dt><a href="http://www.gnu.org/software/dejagnu/">DejaGNU</a></dt>
63<dd>The Feature and Regressions tests are organized and run by DejaGNU.</dd>
64<dt><a href="http://expect.nist.gov/">Expect</a></dt>
65<dd>Expect is required by DejaGNU.</dd>
66<dt><a href="http://www.tcl.tk/software/tcltk/">tcl</a></dt>
67<dd>Tcl is required by DejaGNU. </dd>
68
69<dt><a href="http://www.netlib.org/f2c">F2C</a></dt>
70<dd>For now, LLVM does not have a Fortran front-end, but using F2C, we can run
71Fortran benchmarks. F2C support must be enabled via <tt>configure</tt> if not
72installed in a standard place. F2C requires three items: the <tt>f2c</tt>
73executable, <tt>f2c.h</tt> to compile the generated code, and <tt>libf2c.a</tt>
74to link generated code. By default, given an F2C directory <tt>$DIR</tt>, the
75configure script will search <tt>$DIR/bin</tt> for <tt>f2c</tt>,
76<tt>$DIR/include</tt> for <tt>f2c.h</tt>, and <tt>$DIR/lib</tt> for
77<tt>libf2c.a</tt>. The default <tt>$DIR</tt> values are: <tt>/usr</tt>,
78<tt>/usr/local</tt>, <tt>/sw</tt>, and <tt>/opt</tt>. If you installed F2C in a
79different location, you must tell <tt>configure</tt>:
80
81<ul>
82<li><tt>./configure --with-f2c=$DIR</tt><br>
83This will specify a new <tt>$DIR</tt> for the above-described search
84process. This will only work if the binary, header, and library are in their
85respective subdirectories of <tt>$DIR</tt>.</li>
86
87<li><tt>./configure --with-f2c-bin=/binary/path --with-f2c-inc=/include/path
88--with-f2c-lib=/lib/path</tt><br>
89This allows you to specify the F2C components separately. Note: if you choose
90this route, you MUST specify all three components, and you need to only specify
91<em>directories</em> where the files are located; do NOT include the
92filenames themselves on the <tt>configure</tt> line.</li>
93</ul></dd>
94</dl>
95
96<p>Darwin (Mac OS X) developers can simplify the installation of Expect and tcl
97by using fink. <tt>fink install expect</tt> will install both. Alternatively,
98Darwinports users can use <tt>sudo port install expect</tt> to install Expect
99and tcl.</p>
100
101</div>
102
103<!--=========================================================================-->
104<div class="doc_section"><a name="quick">Quick Start</a></div>
105<!--=========================================================================-->
106
107<div class="doc_text">
108
109 <p>The tests are located in two separate Subversion modules. The basic feature
110 and regression tests are in the main "llvm" module under the directory
111 <tt>llvm/test</tt>. A more comprehensive test suite that includes whole
112programs in C and C++ is in the <tt>test-suite</tt> module. This module should
113be checked out to the <tt>llvm/projects</tt> directory as llvm-test (for
114historical purpose). When you <tt>configure</tt> the <tt>llvm</tt> module,
115the <tt>llvm-test</tt> directory will be automatically configured.
116Alternatively, you can configure the <tt>test-suite</tt> module manually.</p>
117<p>To run all of the simple tests in LLVM using DejaGNU, use the master Makefile
118 in the <tt>llvm/test</tt> directory:</p>
119<pre>
120% gmake -C llvm/test
121</pre>
122or<br>
123<pre>
124% gmake check
125</pre>
126
127<p>To run only a subdirectory of tests in llvm/test using DejaGNU (ie.
128Regression/Transforms), just set the TESTSUITE variable to the path of the
129subdirectory (relative to <tt>llvm/test</tt>):</p>
130<pre>
131% gmake -C llvm/test TESTSUITE=Regression/Transforms
132</pre>
133
134<p><b>Note: If you are running the tests with <tt>objdir != subdir</tt>, you
135must have run the complete testsuite before you can specify a
136subdirectory.</b></p>
137
138<p>To run the comprehensive test suite (tests that compile and execute whole
139programs), run the <tt>llvm-test</tt> tests:</p>
140
141<pre>
142% cd llvm/projects
143% svn co http://llvm.org/svn/llvm-project/test-suite/trunk llvm-test
144% cd llvm-test
145% ./configure --with-llvmsrc=$LLVM_SRC_ROOT --with-llvmobj=$LLVM_OBJ_ROOT
146% gmake
147</pre>
148
149</div>
150
151<!--=========================================================================-->
152<div class="doc_section"><a name="org">LLVM Test Suite Organization</a></div>
153<!--=========================================================================-->
154
155<div class="doc_text">
156
157<p>The LLVM test suite contains two major categories of tests: code
158fragments and whole programs. Code fragments are in the <tt>llvm</tt> module
159under the <tt>llvm/test</tt> directory. The whole programs
160test suite is in the <tt>llvm-test</tt> module under the main directory.</p>
161
162</div>
163
164<!-- _______________________________________________________________________ -->
165<div class="doc_subsection"><a name="codefragments">Code Fragments</a></div>
166<!-- _______________________________________________________________________ -->
167
168<div class="doc_text">
169
170<p>Code fragments are small pieces of code that test a specific feature of LLVM
171or trigger a specific bug in LLVM. They are usually written in LLVM assembly
172language, but can be written in other languages if the test targets a particular
173language front end.</p>
174
175<p>Code fragments are not complete programs, and they are never executed to
176determine correct behavior.</p>
177
178<p>These code fragment tests are located in the <tt>llvm/test/Features</tt> and
179<tt>llvm/test/Regression</tt> directories.</p>
180
181</div>
182
183<!-- _______________________________________________________________________ -->
184<div class="doc_subsection"><a name="wholeprograms">Whole Programs</a></div>
185<!-- _______________________________________________________________________ -->
186
187<div class="doc_text">
188
189<p>Whole Programs are pieces of code which can be compiled and linked into a
190stand-alone program that can be executed. These programs are generally written
191in high level languages such as C or C++, but sometimes they are written
192straight in LLVM assembly.</p>
193
194<p>These programs are compiled and then executed using several different
195methods (native compiler, LLVM C backend, LLVM JIT, LLVM native code generation,
196etc). The output of these programs is compared to ensure that LLVM is compiling
197the program correctly.</p>
198
199<p>In addition to compiling and executing programs, whole program tests serve as
200a way of benchmarking LLVM performance, both in terms of the efficiency of the
201programs generated as well as the speed with which LLVM compiles, optimizes, and
202generates code.</p>
203
204<p>All "whole program" tests are located in the <tt>test-suite</tt> Subversion
205module.</p>
206
207</div>
208
209<!--=========================================================================-->
210<div class="doc_section"><a name="tree">LLVM Test Suite Tree</a></div>
211<!--=========================================================================-->
212
213<div class="doc_text">
214
215<p>Each type of test in the LLVM test suite has its own directory. The major
216subtrees of the test suite directory tree are as follows:</p>
217
218<ul>
219 <li><tt>llvm/test</tt>
220 <p>This directory contains a large array of small tests
221 that exercise various features of LLVM and to ensure that regressions do not
222 occur. The directory is broken into several sub-directories, each focused on
223 a particular area of LLVM. A few of the important ones are:<ul>
224 <li><tt>Analysis</tt>: checks Analysis passes.</li>
225 <li><tt>Archive</tt>: checks the Archive library.</li>
226 <li><tt>Assembler</tt>: checks Assembly reader/writer functionality.</li>
227 <li><tt>Bitcode</tt>: checks Bitcode reader/writer functionality.</li>
228 <li><tt>CodeGen</tt>: checks code generation and each target.</li>
229 <li><tt>Features</tt>: checks various features of the LLVM language.</li>
230 <li><tt>Linker</tt>: tests bitcode linking.</li>
231 <li><tt>Transforms</tt>: tests each of the scalar, IPO, and utility
232 transforms to ensure they make the right transformations.</li>
233 <li><tt>Verifier</tt>: tests the IR verifier.</li>
234 </ul></p>
235 <p>Typically when a bug is found in LLVM, a regression test containing
236 just enough code to reproduce the problem should be written and placed
237 somewhere underneath this directory. In most cases, this will be a small
238 piece of LLVM assembly language code, often distilled from an actual
239 application or benchmark.</p></li>
240
241<li><tt>test-suite</tt>
242<p>The <tt>test-suite</tt> module contains programs that can be compiled
243with LLVM and executed. These programs are compiled using the native compiler
244and various LLVM backends. The output from the program compiled with the
245native compiler is assumed correct; the results from the other programs are
246compared to the native program output and pass if they match.</p>
247
248<p>In addition for testing correctness, the <tt>llvm-test</tt> directory also
249performs timing tests of various LLVM optimizations. It also records
250compilation times for the compilers and the JIT. This information can be
251used to compare the effectiveness of LLVM's optimizations and code
252generation.</p></li>
253
254<li><tt>llvm-test/SingleSource</tt>
255<p>The SingleSource directory contains test programs that are only a single
256source file in size. These are usually small benchmark programs or small
257programs that calculate a particular value. Several such programs are grouped
258together in each directory.</p></li>
259
260<li><tt>llvm-test/MultiSource</tt>
261<p>The MultiSource directory contains subdirectories which contain entire
262programs with multiple source files. Large benchmarks and whole applications
263go here.</p></li>
264
265<li><tt>llvm-test/External</tt>
266<p>The External directory contains Makefiles for building code that is external
267to (i.e., not distributed with) LLVM. The most prominent members of this
268directory are the SPEC 95 and SPEC 2000 benchmark suites. The presence and
269location of these external programs is configured by the llvm-test
270<tt>configure</tt> script.</p></li>
271
272</ul>
273
274</div>
275<!--=========================================================================-->
276<div class="doc_section"><a name="dgstructure">DejaGNU Structure</a></div>
277<!--=========================================================================-->
278<div class="doc_text">
279 <p>The LLVM test suite is partially driven by DejaGNU and partially driven by
280 GNU Make. Specifically, the Features and Regression tests are all driven by
281 DejaGNU. The <tt>llvm-test</tt> module is currently driven by a set of
282 Makefiles.</p>
283
284 <p>The DejaGNU structure is very simple, but does require some information to
285 be set. This information is gathered via <tt>configure</tt> and is written
286 to a file, <tt>site.exp</tt> in <tt>llvm/test</tt>. The <tt>llvm/test</tt>
287 Makefile does this work for you.</p>
288
289 <p>In order for DejaGNU to work, each directory of tests must have a
290 <tt>dg.exp</tt> file. DejaGNU looks for this file to determine how to run the
291 tests. This file is just a Tcl script and it can do anything you want, but
292 we've standardized it for the LLVM regression tests. It simply loads a Tcl
293 library (<tt>test/lib/llvm.exp</tt>) and calls the <tt>llvm_runtests</tt>
294 function defined in that library with a list of file names to run. The names
295 are obtained by using Tcl's glob command. Any directory that contains only
296 directories does not need the <tt>dg.exp</tt> file.</p>
297
298 <p>The <tt>llvm-runtests</tt> function lookas at each file that is passed to
299 it and gathers any lines together that match "RUN:". This are the "RUN" lines
300 that specify how the test is to be run. So, each test script must contain
301 RUN lines if it is to do anything. If there are no RUN lines, the
302 <tt>llvm-runtests</tt> function will issue an error and the test will
303 fail.</p>
304
305 <p>RUN lines are specified in the comments of the test program using the
306 keyword <tt>RUN</tt> followed by a colon, and lastly the command (pipeline)
307 to execute. Together, these lines form the "script" that
308 <tt>llvm-runtests</tt> executes to run the test case. The syntax of the
309 RUN lines is similar to a shell's syntax for pipelines including I/O
310 redirection and variable substitution. However, even though these lines
311 may <i>look</i> like a shell script, they are not. RUN lines are interpreted
312 directly by the Tcl <tt>exec</tt> command. They are never executed by a
313 shell. Consequently the syntax differs from normal shell script syntax in a
314 few ways. You can specify as many RUN lines as needed.</p>
315
316 <p>Each RUN line is executed on its own, distinct from other lines unless
317 its last character is <tt>\</tt>. This continuation character causes the RUN
318 line to be concatenated with the next one. In this way you can build up long
319 pipelines of commands without making huge line lengths. The lines ending in
320 <tt>\</tt> are concatenated until a RUN line that doesn't end in <tt>\</tt> is
321 found. This concatenated set or RUN lines then constitutes one execution.
322 Tcl will substitute variables and arrange for the pipeline to be executed. If
323 any process in the pipeline fails, the entire line (and test case) fails too.
324 </p>
325
326 <p> Below is an example of legal RUN lines in a <tt>.ll</tt> file:</p>
327 <pre>
328 ; RUN: llvm-as &lt; %s | llvm-dis &gt; %t1
329 ; RUN: llvm-dis &lt; %s.bc-13 &gt; %t2
330 ; RUN: diff %t1 %t2
331 </pre>
332
333 <p>As with a Unix shell, the RUN: lines permit pipelines and I/O redirection
334 to be used. However, the usage is slightly different than for Bash. To check
335 what's legal, see the documentation for the
336 <a href="http://www.tcl.tk/man/tcl8.5/TclCmd/exec.htm#M2">Tcl exec</a>
337 command and the
338 <a href="http://www.tcl.tk/man/tcl8.5/tutorial/Tcl26.html">tutorial</a>.
339 The major differences are:</p>
340 <ul>
341 <li>You can't do <tt>2&gt;&amp;1</tt>. That will cause Tcl to write to a
342 file named <tt>&amp;1</tt>. Usually this is done to get stderr to go through
343 a pipe. You can do that in tcl with <tt>|&amp;</tt> so replace this idiom:
344 <tt>... 2&gt;&amp;1 | grep</tt> with <tt>... |&amp; grep</tt></li>
345 <li>You can only redirect to a file, not to another descriptor and not from
346 a here document.</li>
347 <li>tcl supports redirecting to open files with the @ syntax but you
348 shouldn't use that here.</li>
349 </ul>
350
351 <p>There are some quoting rules that you must pay attention to when writing
352 your RUN lines. In general nothing needs to be quoted. Tcl won't strip off any
353 ' or " so they will get passed to the invoked program. For example:</p>
354 <pre>
355 ... | grep 'find this string'
356 </pre>
357 <p>This will fail because the ' characters are passed to grep. This would
358 instruction grep to look for <tt>'find</tt> in the files <tt>this</tt> and
359 <tt>string'</tt>. To avoid this use curly braces to tell Tcl that it should
360 treat everything enclosed as one value. So our example would become:</p>
361 <pre>
362 ... | grep {find this string}
363 </pre>
364 <p>Additionally, the characters <tt>[</tt> and <tt>]</tt> are treated
365 specially by Tcl. They tell Tcl to interpret the content as a command to
366 execute. Since these characters are often used in regular expressions this can
367 have disastrous results and cause the entire test run in a directory to fail.
368 For example, a common idiom is to look for some basicblock number:</p>
369 <pre>
370 ... | grep bb[2-8]
371 </pre>
372 <p>This, however, will cause Tcl to fail because its going to try to execute
373 a program named "2-8". Instead, what you want is this:</p>
374 <pre>
375 ... | grep {bb\[2-8\]}
376 </pre>
377 <p>Finally, if you need to pass the <tt>\</tt> character down to a program,
378 then it must be doubled. This is another Tcl special character. So, suppose
379 you had:
380 <pre>
381 ... | grep 'i32\*'
382 </pre>
383 <p>This will fail to match what you want (a pointer to i32). First, the
384 <tt>'</tt> do not get stripped off. Second, the <tt>\</tt> gets stripped off
385 by Tcl so what grep sees is: <tt>'i32*'</tt>. That's not likely to match
386 anything. To resolve this you must use <tt>\\</tt> and the <tt>{}</tt>, like
387 this:</p>
388 <pre>
389 ... | grep {i32\\*}
390 </pre>
391
392</div>
393
394<!-- _______________________________________________________________________ -->
395<div class="doc_subsection"><a name="dgvars">Vars And Substitutions</a></div>
396<div class="doc_text">
397 <p>With a RUN line there are a number of substitutions that are permitted. In
398 general, any Tcl variable that is available in the <tt>substitute</tt>
399 function (in <tt>test/lib/llvm.exp</tt>) can be substituted into a RUN line.
400 To make a substitution just write the variable's name preceded by a $.
401 Additionally, for compatibility reasons with previous versions of the test
402 library, certain names can be accessed with an alternate syntax: a % prefix.
403 These alternates are deprecated and may go away in a future version.
404 </p>
405 Here are the available variable names. The alternate syntax is listed in
406 parentheses.</p>
407 <dl style="margin-left: 25px">
408 <dt><b>$test</b> (%s)</dt>
409 <dd>The full path to the test case's source. This is suitable for passing
410 on the command line as the input to an llvm tool.</dd>
411 <dt><b>$srcdir</b></dt>
412 <dd>The source directory from where the "<tt>make check</tt>" was run.</dd>
413 <dt><b>objdir</b></dt>
414 <dd>The object directory that corresponds to the </tt>$srcdir</tt>.</dd>
415 <dt><b>subdir</b></dt>
416 <dd>A partial path from the <tt>test</tt> directory that contains the
417 sub-directory that contains the test source being executed.</dd>
418 <dt><b>srcroot</b></dt>
419 <dd>The root directory of the LLVM src tree.</dd>
420 <dt><b>objroot</b></dt>
421 <dd>The root directory of the LLVM object tree. This could be the same
422 as the srcroot.</dd>
423 <dt><b>path</b><dt>
424 <dd>The path to the directory that contains the test case source. This is
425 for locating any supporting files that are not generated by the test, but
426 used by the test.</dd>
427 <dt><b>tmp</b></dt>
428 <dd>The path to a temporary file name that could be used for this test case.
429 The file name won't conflict with other test cases. You can append to it if
430 you need multiple temporaries. This is useful as the destination of some
431 redirected output.</dd>
432 <dt><b>llvmlibsdir</b> (%llvmlibsdir)</dt>
433 <dd>The directory where the LLVM libraries are located.</dd>
434 <dt><b>target_triplet</b> (%target_triplet)</dt>
435 <dd>The target triplet that corresponds to the current host machine (the one
436 running the test cases). This should probably be called "host".<dd>
437 <dt><b>prcontext</b> (%prcontext)</dt>
438 <dd>Path to the prcontext tcl script that prints some context around a
439 line that matches a pattern. This isn't strictly necessary as the test suite
440 is run with its PATH altered to include the test/Scripts directory where
441 the prcontext script is located. Note that this script is similar to
442 <tt>grep -C</tt> but you should use the <tt>prcontext</tt> script because
443 not all platforms support <tt>grep -C</tt>.</dd>
444 <dt><b>llvmgcc</b> (%llvmgcc)</dt>
445 <dd>The full path to the <tt>llvm-gcc</tt> executable as specified in the
446 configured LLVM environment</dd>
447 <dt><b>llvmgxx</b> (%llvmgxx)</dt>
448 <dd>The full path to the <tt>llvm-gxx</tt> executable as specified in the
449 configured LLVM environment</dd>
450 <dt><b>llvmgcc_version</b> (%llvmgcc_version)</dt>
451 <dd>The full version number of the <tt>llvm-gcc</tt> executable.</dd>
452 <dt><b>llvmgccmajvers</b> (%llvmgccmajvers)</dt>
453 <dd>The major version number of the <tt>llvm-gcc</tt> executable.</dd>
454 <dt><b>gccpath</b></dt>
455 <dd>The full path to the C compiler used to <i>build </i> LLVM. Note that
456 this might not be gcc.</dd>
457 <dt><b>gxxpath</b></dt>
458 <dd>The full path to the C++ compiler used to <i>build </i> LLVM. Note that
459 this might not be g++.</dd>
460 <dt><b>compile_c</b> (%compile_c)</dt>
461 <dd>The full command line used to compile LLVM C source code. This has all
462 the configured -I, -D and optimization options.</dd>
463 <dt><b>compile_cxx</b> (%compile_cxx)</dt>
464 <dd>The full command used to compile LLVM C++ source code. This has
465 all the configured -I, -D and optimization options.</dd>
466 <dt><b>link</b> (%link)</dt>
467 <dd>This full link command used to link LLVM executables. This has all the
468 configured -I, -L and -l options.</dd>
469 <dt><b>shlibext</b> (%shlibext)</dt>
470 <dd>The suffix for the host platforms share library (dll) files. This
471 includes the period as the first character.</dd>
472 </dl>
473 <p>To add more variables, two things need to be changed. First, add a line in
474 the <tt>test/Makefile</tt> that creates the <tt>site.exp</tt> file. This will
475 "set" the variable as a global in the site.exp file. Second, in the
476 <tt>test/lib/llvm.exp</tt> file, in the substitute proc, add the variable name
477 to the list of "global" declarations at the beginning of the proc. That's it,
478 the variable can then be used in test scripts.</p>
479</div>
480
481<!-- _______________________________________________________________________ -->
482<div class="doc_subsection"><a name="dgfeatures">Other Features</a></div>
483<div class="doc_text">
484 <p>To make RUN line writing easier, there are several shell scripts located
485 in the <tt>llvm/test/Scripts</tt> directory. For example:</p>
486 <dl>
487 <dt><b>ignore</b></dt>
488 <dd>This script runs its arguments and then always returns 0. This is useful
489 in cases where the test needs to cause a tool to generate an error (e.g. to
490 check the error output). However, any program in a pipeline that returns a
491 non-zero result will cause the test to fail. This script overcomes that
492 issue and nicely documents that the test case is purposefully ignoring the
493 result code of the tool</dd>
494 <dt><b>not</b></dt>
495 <dd>This script runs its arguments and then inverts the result code from
496 it. Zero result codes become 1. Non-zero result codes become 0. This is
497 useful to invert the result of a grep. For example "not grep X" means
498 succeed only if you don't find X in the input.</dd>
499 </dl>
500
501 <p>Sometimes it is necessary to mark a test case as "expected fail" or XFAIL.
502 You can easily mark a test as XFAIL just by including <tt>XFAIL: </tt> on a
503 line near the top of the file. This signals that the test case should succeed
504 if the test fails. Such test cases are counted separately by DejaGnu. To
505 specify an expected fail, use the XFAIL keyword in the comments of the test
506 program followed by a colon and one or more regular expressions (separated by
507 a comma). The regular expressions allow you to XFAIL the test conditionally
508 by host platform. The regular expressions following the : are matched against
509 the target triplet or llvmgcc version number for the host machine. If there is
510 a match, the test is expected to fail. If not, the test is expected to
511 succeed. To XFAIL everywhere just specify <tt>XFAIL: *</tt>. When matching
512 the llvm-gcc version, you can specify the major (e.g. 3) or full version
513 (i.e. 3.4) number. Here is an example of an <tt>XFAIL</tt> line:</p>
514 <pre>
515 ; XFAIL: darwin,sun,llvmgcc4
516 </pre>
517
518 <p>To make the output more useful, the <tt>llvm_runtest</tt> function wil
519 scan the lines of the test case for ones that contain a pattern that matches
520 PR[0-9]+. This is the syntax for specifying a PR (Problem Report) number that
521 is related to the test case. The numer after "PR" specifies the LLVM bugzilla
522 number. When a PR number is specified, it will be used in the pass/fail
523 reporting. This is useful to quickly get some context when a test fails.</p>
524
525 <p>Finally, any line that contains "END." will cause the special
526 interpretation of lines to terminate. This is generally done right after the
527 last RUN: line. This has two side effects: (a) it prevents special
528 interpretation of lines that are part of the test program, not the
529 instructions to the test case, and (b) it speeds things up for really big test
530 cases by avoiding interpretation of the remainder of the file.</p>
531
532</div>
533
534<!--=========================================================================-->
535<div class="doc_section"><a name="progstructure"><tt>llvm-test</tt>
536Structure</a></div>
537<!--=========================================================================-->
538
539<div class="doc_text">
540
541<p>As mentioned previously, the <tt>llvm-test</tt> module provides three types
542of tests: MultiSource, SingleSource, and External. Each tree is then subdivided
543into several categories, including applications, benchmarks, regression tests,
544code that is strange grammatically, etc. These organizations should be
545relatively self explanatory.</p>
546
547<p>In addition to the regular "whole program" tests, the <tt>llvm-test</tt>
548module also provides a mechanism for compiling the programs in different ways.
549If the variable TEST is defined on the gmake command line, the test system will
550include a Makefile named <tt>TEST.&lt;value of TEST variable&gt;.Makefile</tt>.
551This Makefile can modify build rules to yield different results.</p>
552
553<p>For example, the LLVM nightly tester uses <tt>TEST.nightly.Makefile</tt> to
554create the nightly test reports. To run the nightly tests, run <tt>gmake
555TEST=nightly</tt>.</p>
556
557<p>There are several TEST Makefiles available in the tree. Some of them are
558designed for internal LLVM research and will not work outside of the LLVM
559research group. They may still be valuable, however, as a guide to writing your
560own TEST Makefile for any optimization or analysis passes that you develop with
561LLVM.</p>
562
563<p>Note, when configuring the <tt>llvm-test</tt> module, you might want to
564specify the following configuration options:</p>
565<dl>
566 <dt><i>--enable-spec2000</i>
567 <dt><i>--enable-spec2000=&lt;<tt>directory</tt>&gt;</i>
568 <dd>
569 Enable the use of SPEC2000 when testing LLVM. This is disabled by default
570 (unless <tt>configure</tt> finds SPEC2000 installed). By specifying
571 <tt>directory</tt>, you can tell configure where to find the SPEC2000
572 benchmarks. If <tt>directory</tt> is left unspecified, <tt>configure</tt>
573 uses the default value
574 <tt>/home/vadve/shared/benchmarks/speccpu2000/benchspec</tt>.
575 <p>
576 <dt><i>--enable-spec95</i>
577 <dt><i>--enable-spec95=&lt;<tt>directory</tt>&gt;</i>
578 <dd>
579 Enable the use of SPEC95 when testing LLVM. It is similar to the
580 <i>--enable-spec2000</i> option.
581 <p>
582 <dt><i>--enable-povray</i>
583 <dt><i>--enable-povray=&lt;<tt>directory</tt>&gt;</i>
584 <dd>
585 Enable the use of Povray as an external test. Versions of Povray written
586 in C should work. This option is similar to the <i>--enable-spec2000</i>
587 option.
588</dl>
589</div>
590
591<!--=========================================================================-->
592<div class="doc_section"><a name="run">Running the LLVM Tests</a></div>
593<!--=========================================================================-->
594
595<div class="doc_text">
596
597<p>First, all tests are executed within the LLVM object directory tree. They
598<i>are not</i> executed inside of the LLVM source tree. This is because the
599test suite creates temporary files during execution.</p>
600
601<p>The master Makefile in llvm/test is capable of running only the DejaGNU
602driven tests. By default, it will run all of these tests.</p>
603
604<p>To run only the DejaGNU driven tests, run <tt>gmake</tt> at the
605command line in <tt>llvm/test</tt>. To run a specific directory of tests, use
606the TESTSUITE variable.
607</p>
608
609<p>For example, to run the Regression tests, type
610<tt>gmake TESTSUITE=Regression</tt> in <tt>llvm/tests</tt>.</p>
611
612<p>Note that there are no Makefiles in <tt>llvm/test/Features</tt> and
613<tt>llvm/test/Regression</tt>. You must use DejaGNU from the <tt>llvm/test</tt>
614directory to run them.</p>
615
616<p>To run the <tt>llvm-test</tt> suite, you need to use the following steps:
617</p>
618<ol>
619 <li>cd into the llvm/projects directory</li>
620 <li>check out the <tt>test-suite</tt> module with:<br/>
621 <tt>svn co http://llvm.org/svn/llvm-project/test-suite/trunk llvm-test<br/>
622 This will get the test suite into <tt>llvm/projects/llvm-test</tt></li>
623 <li>configure the test suite. You can do this one of two ways:
624 <ol>
625 <li>Use the regular llvm configure:<br/>
626 <tt>cd $LLVM_OBJ_ROOT ; $LLVM_SRC_ROOT/configure</tt><br/>
627 This will ensure that the <tt>projects/llvm-test</tt> directory is also
628 properly configured.</li>
629 <li>Use the <tt>configure</tt> script found in the <tt>llvm-test</tt> source
630 directory:<br/>
631 <tt>$LLVM_SRC_ROOT/projects/llvm-test/configure
632 --with-llvmsrc=$LLVM_SRC_ROOT --with-llvmobj=$LLVM_OBJ_ROOT</tt>
633 </li>
634 </ol>
635 <li>gmake</li>
636</ol>
637<p>Note that the second and third steps only need to be done once. After you
638have the suite checked out and configured, you don't need to do it again (unless
639the test code or configure script changes).</p>
640
641<p>To make a specialized test (use one of the
642<tt>llvm-test/TEST.&lt;type&gt;.Makefile</tt>s), just run:<br/>
643<tt>gmake TEST=&lt;type&gt; test</tt><br/>For example, you could run the
644nightly tester tests using the following commands:</p>
645
646<pre>
647 % cd llvm/projects/llvm-test
648 % gmake TEST=nightly test
649</pre>
650
651<p>Regardless of which test you're running, the results are printed on standard
652output and standard error. You can redirect these results to a file if you
653choose.</p>
654
655<p>Some tests are known to fail. Some are bugs that we have not fixed yet;
656others are features that we haven't added yet (or may never add). In DejaGNU,
657the result for such tests will be XFAIL (eXpected FAILure). In this way, you
658can tell the difference between an expected and unexpected failure.</p>
659
660<p>The tests in <tt>llvm-test</tt> have no such feature at this time. If the
661test passes, only warnings and other miscellaneous output will be generated. If
662a test fails, a large &lt;program&gt; FAILED message will be displayed. This
663will help you separate benign warnings from actual test failures.</p>
664
665</div>
666
667<!-- _______________________________________________________________________ -->
668<div class="doc_subsection">
669<a name="customtest">Writing custom tests for llvm-test</a></div>
670<!-- _______________________________________________________________________ -->
671
672<div class="doc_text">
673
674<p>Assuming you can run llvm-test, (e.g. "<tt>gmake TEST=nightly report</tt>"
675should work), it is really easy to run optimizations or code generator
676components against every program in the tree, collecting statistics or running
677custom checks for correctness. At base, this is how the nightly tester works,
678it's just one example of a general framework.</p>
679
680<p>Lets say that you have an LLVM optimization pass, and you want to see how
681many times it triggers. First thing you should do is add an LLVM
682<a href="ProgrammersManual.html#Statistic">statistic</a> to your pass, which
683will tally counts of things you care about.</p>
684
685<p>Following this, you can set up a test and a report that collects these and
686formats them for easy viewing. This consists of two files, an
687"<tt>llvm-test/TEST.XXX.Makefile</tt>" fragment (where XXX is the name of your
688test) and an "<tt>llvm-test/TEST.XXX.report</tt>" file that indicates how to
689format the output into a table. There are many example reports of various
690levels of sophistication included with llvm-test, and the framework is very
691general.</p>
692
693<p>If you are interested in testing an optimization pass, check out the
694"libcalls" test as an example. It can be run like this:<p>
695
696<div class="doc_code">
697<pre>
698% cd llvm/projects/llvm-test/MultiSource/Benchmarks # or some other level
699% make TEST=libcalls report
700</pre>
701</div>
702
703<p>This will do a bunch of stuff, then eventually print a table like this:</p>
704
705<div class="doc_code">
706<pre>
707Name | total | #exit |
708...
709FreeBench/analyzer/analyzer | 51 | 6 |
710FreeBench/fourinarow/fourinarow | 1 | 1 |
711FreeBench/neural/neural | 19 | 9 |
712FreeBench/pifft/pifft | 5 | 3 |
713MallocBench/cfrac/cfrac | 1 | * |
714MallocBench/espresso/espresso | 52 | 12 |
715MallocBench/gs/gs | 4 | * |
716Prolangs-C/TimberWolfMC/timberwolfmc | 302 | * |
717Prolangs-C/agrep/agrep | 33 | 12 |
718Prolangs-C/allroots/allroots | * | * |
719Prolangs-C/assembler/assembler | 47 | * |
720Prolangs-C/bison/mybison | 74 | * |
721...
722</pre>
723</div>
724
725<p>This basically is grepping the -stats output and displaying it in a table.
726You can also use the "TEST=libcalls report.html" target to get the table in HTML
727form, similarly for report.csv and report.tex.</p>
728
729<p>The source for this is in llvm-test/TEST.libcalls.*. The format is pretty
730simple: the Makefile indicates how to run the test (in this case,
731"<tt>opt -simplify-libcalls -stats</tt>"), and the report contains one line for
732each column of the output. The first value is the header for the column and the
733second is the regex to grep the output of the command for. There are lots of
734example reports that can do fancy stuff.</p>
735
736</div>
737
738
739<!--=========================================================================-->
740<div class="doc_section"><a name="nightly">Running the nightly tester</a></div>
741<!--=========================================================================-->
742
743<div class="doc_text">
744
745<p>
746The <a href="http://llvm.org/nightlytest/">LLVM Nightly Testers</a>
747automatically check out an LLVM tree, build it, run the "nightly"
748program test (described above), run all of the feature and regression tests,
749delete the checked out tree, and then submit the results to
750<a href="http://llvm.org/nightlytest/">http://llvm.org/nightlytest/</a>.
751After test results are submitted to
752<a href="http://llvm.org/nightlytest/">http://llvm.org/nightlytest/</a>,
753they are processed and displayed on the tests page. An email to
754<a href="http://lists.cs.uiuc.edu/pipermail/llvm-testresults/">
755llvm-testresults@cs.uiuc.edu</a> summarizing the results is also generated.
756This testing scheme is designed to ensure that programs don't break as well
757as keep track of LLVM's progress over time.</p>
758
759<p>If you'd like to set up an instance of the nightly tester to run on your
760machine, take a look at the comments at the top of the
761<tt>utils/NewNightlyTest.pl</tt> file. If you decide to set up a nightly tester
762please choose a unique nickname and invoke <tt>utils/NewNightlyTest.pl</tt>
763with the "-nickname [yournickname]" command line option.
764
765<p>You can create a shell script to encapsulate the running of the script.
766The optimized x86 Linux nightly test is run from just such a script:</p>
767
768<div class="doc_code">
769<pre>
770#!/bin/bash
771BASE=/proj/work/llvm/nightlytest
772export BUILDDIR=$BASE/build
773export WEBDIR=$BASE/testresults
774export LLVMGCCDIR=/proj/work/llvm/cfrontend/install
775export PATH=/proj/install/bin:$LLVMGCCDIR/bin:$PATH
776export LD_LIBRARY_PATH=/proj/install/lib
777cd $BASE
778cp /proj/work/llvm/llvm/utils/NewNightlyTest.pl .
779nice ./NewNightlyTest.pl -nice -release -verbose -parallel -enable-linscan \
780 -nickname NightlyTester -noexternals &gt; output.log 2&gt;&amp;1
781</pre>
782</div>
783
784<p>It is also possible to specify the the location your nightly test results
785are submitted. You can do this by passing the command line option
786"-submit-server [server_address]" and "-submit-script [script_on_server]" to
787<tt>utils/NewNightlyTest.pl</tt>. For example, to submit to the llvm.org
788nightly test results page, you would invoke the nightly test script with
789"-submit-server llvm.org -submit-script /nightlytest/NightlyTestAccept.cgi".
790If these options are not specified, the nightly test script sends the results
791to the llvm.org nightly test results page.</p>
792
793<p>Take a look at the <tt>NewNightlyTest.pl</tt> file to see what all of the
794flags and strings do. If you start running the nightly tests, please let us
795know. Thanks!</p>
796
797</div>
798
799<!-- *********************************************************************** -->
800
801<hr>
802<address>
803 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
804 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
805 <a href="http://validator.w3.org/check/referer"><img
806 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
807
808 John T. Criswell, Reid Spencer, and Tanya Lattner<br>
809 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br/>
810 Last modified: $Date$
811</address>
812</body>
813</html>