docs/TestingGuide.html

   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 Testing Infrastructure Guide</title>
   6   <link rel="stylesheet" href="llvm.css" type="text/css">
   7 </head>
   8 <body>
   9
  10 <div class="doc_title">
  11   LLVM Testing Infrastructure 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="#org">LLVM testing infrastructure organization</a>
  18     <ul>
  19       <li><a href="#dejagnu">DejaGNU tests</a></li>
  20       <li><a href="#testsuite">Test suite</a></li>
  21     </ul>
  22   </li>
  23   <li><a href="#quick">Quick start</a>
  24     <ul>
  25       <li><a href="#quickdejagnu">DejaGNU tests</a></li>
  26       <li><a href="#quicktestsuite">Test suite</a></li>
  27    </ul>
  28   </li>
  29   <li><a href="#dgstructure">DejaGNU structure</a>
  30     <ul>
  31       <li><a href="#dgcustom">Writing new DejaGNU tests</a></li>
  32       <li><a href="#FileCheck">The FileCheck utility</a></li>
  33       <li><a href="#dgvars">Variables and substitutions</a></li>
  34       <li><a href="#dgfeatures">Other features</a></li>
  35    </ul>
  36   </li>
  37   <li><a href="#testsuitestructure">Test suite structure</a></li>
  38   <li><a href="#testsuiterun">Running the test suite</a>
  39     <ul>
  40       <li><a href="#testsuiteexternal">Configuring External Tests</a></li>
  41       <li><a href="#testsuitetests">Running different tests</a></li>
  42       <li><a href="#testsuiteoutput">Generating test output</a></li>
  43       <li><a href="#testsuitecustom">Writing custom tests for llvm-test</a></li>
  44    </ul>
  45   </li>
  46   <li><a href="#nightly">Running the nightly tester</a></li>
  47 </ol>
  48
  49 <div class="doc_author">
  50   <p>Written by John T. Criswell, <a
  51   href="http://llvm.x10sys.com/rspencer">Reid Spencer</a>, and Tanya Lattner</p>
  52 </div>
  53
  54 <!--=========================================================================-->
  55 <div class="doc_section"><a name="overview">Overview</a></div>
  56 <!--=========================================================================-->
  57
  58 <div class="doc_text">
  59
  60 <p>This document is the reference manual for the LLVM testing infrastructure. It documents
  61 the structure of the LLVM testing infrastructure, the tools needed to use it,
  62 and how to add and run tests.</p>
  63
  64 </div>
  65
  66 <!--=========================================================================-->
  67 <div class="doc_section"><a name="requirements">Requirements</a></div>
  68 <!--=========================================================================-->
  69
  70 <div class="doc_text">
  71
  72 <p>In order to use the LLVM testing infrastructure, you will need all of the software
  73 required to build LLVM, plus the following:</p>
  74
  75 <dl>
  76 <dt><a href="http://www.gnu.org/software/dejagnu/">DejaGNU</a></dt>
  77 <dd>The Feature and Regressions tests are organized and run by DejaGNU.</dd>
  78 <dt><a href="http://expect.nist.gov/">Expect</a></dt>
  79 <dd>Expect is required by DejaGNU.</dd>
  80 <dt><a href="http://www.tcl.tk/software/tcltk/">tcl</a></dt>
  81 <dd>Tcl is required by DejaGNU. </dd>
  82 </dl>
  83
  84 </div>
  85
  86 <!--=========================================================================-->
  87 <div class="doc_section"><a name="org">LLVM testing infrastructure organization</a></div>
  88 <!--=========================================================================-->
  89
  90 <div class="doc_text">
  91
  92 <p>The LLVM testing infrastructure contains two major categories of tests: code
  93 fragments and whole programs. Code fragments are referred to as the "DejaGNU
  94 tests" and are in the <tt>llvm</tt> module in subversion under the
  95 <tt>llvm/test</tt> directory. The whole programs tests are referred to as the
  96 "Test suite" and are in the <tt>test-suite</tt> module in subversion.
  97 </p>
  98
  99 </div>
 100
 101 <!-- _______________________________________________________________________ -->
 102 <div class="doc_subsection"><a name="dejagnu">DejaGNU tests</a></div>
 103 <!-- _______________________________________________________________________ -->
 104
 105 <div class="doc_text">
 106
 107 <p>Code fragments are small pieces of code that test a specific
 108 feature of LLVM or trigger a specific bug in LLVM.  They are usually
 109 written in LLVM assembly language, but can be written in other
 110 languages if the test targets a particular language front end (and the
 111 appropriate <tt>--with-llvmgcc</tt> options were used
 112 at <tt>configure</tt> time of the <tt>llvm</tt> module). These tests
 113 are driven by the DejaGNU testing framework, which is hidden behind a
 114 few simple makefiles.</p>
 115
 116 <p>These code fragments are not complete programs. The code generated
 117 from them is never executed to determine correct behavior.</p>
 118
 119 <p>These code fragment tests are located in the <tt>llvm/test</tt>
 120 directory.</p>
 121
 122 <p>Typically when a bug is found in LLVM, a regression test containing
 123 just enough code to reproduce the problem should be written and placed
 124 somewhere underneath this directory.  In most cases, this will be a small
 125 piece of LLVM assembly language code, often distilled from an actual
 126 application or benchmark.</p>
 127
 128 </div>
 129
 130 <!-- _______________________________________________________________________ -->
 131 <div class="doc_subsection"><a name="testsuite">Test suite</a></div>
 132 <!-- _______________________________________________________________________ -->
 133
 134 <div class="doc_text">
 135
 136 <p>The test suite contains whole programs, which are pieces of
 137 code which can be compiled and linked into a stand-alone program that can be
 138 executed.  These programs are generally written in high level languages such as
 139 C or C++, but sometimes they are written straight in LLVM assembly.</p>
 140
 141 <p>These programs are compiled and then executed using several different
 142 methods (native compiler, LLVM C backend, LLVM JIT, LLVM native code generation,
 143 etc).  The output of these programs is compared to ensure that LLVM is compiling
 144 the program correctly.</p>
 145
 146 <p>In addition to compiling and executing programs, whole program tests serve as
 147 a way of benchmarking LLVM performance, both in terms of the efficiency of the
 148 programs generated as well as the speed with which LLVM compiles, optimizes, and
 149 generates code.</p>
 150
 151 <p>The test-suite is located in the <tt>test-suite</tt> Subversion module.</p>
 152
 153 </div>
 154
 155 <!--=========================================================================-->
 156 <div class="doc_section"><a name="quick">Quick start</a></div>
 157 <!--=========================================================================-->
 158
 159 <div class="doc_text">
 160
 161   <p>The tests are located in two separate Subversion modules. The
 162   DejaGNU tests are in the main "llvm" module under the directory
 163   <tt>llvm/test</tt> (so you get these tests for free with the main llvm tree).
 164   The more comprehensive test suite that includes whole
 165 programs in C and C++ is in the <tt>test-suite</tt> module. This module should
 166 be checked out to the <tt>llvm/projects</tt> directory (don't use another name
 167 then the default "test-suite", for then the test suite will be run every time
 168 you run <tt>make</tt> in the main <tt>llvm</tt> directory).
 169 When you <tt>configure</tt> the <tt>llvm</tt> module,
 170 the <tt>test-suite</tt> directory will be automatically configured.
 171 Alternatively, you can configure the <tt>test-suite</tt> module manually.</p>
 172
 173 <!-- _______________________________________________________________________ -->
 174 <div class="doc_subsection"><a name="quickdejagnu">DejaGNU tests</a></div>
 175 <!-- _______________________________________________________________________ -->
 176 <p>To run all of the simple tests in LLVM using DejaGNU, use the master Makefile
 177  in the <tt>llvm/test</tt> directory:</p>
 178
 179 <div class="doc_code">
 180 <pre>
 181 % gmake -C llvm/test
 182 </pre>
 183 </div>
 184
 185 <p>or</p>
 186
 187 <div class="doc_code">
 188 <pre>
 189 % gmake check
 190 </pre>
 191 </div>
 192
 193 <p>To run only a subdirectory of tests in <tt>llvm/test</tt> using DejaGNU (ie.
 194 Transforms), just set the TESTSUITE variable to the path of the
 195 subdirectory (relative to <tt>llvm/test</tt>):</p>
 196
 197 <div class="doc_code">
 198 <pre>
 199 % gmake TESTSUITE=Transforms check
 200 </pre>
 201 </div>
 202
 203 <p><b>Note: If you are running the tests with <tt>objdir != subdir</tt>, you
 204 must have run the complete testsuite before you can specify a
 205 subdirectory.</b></p>
 206
 207 <p>To run only a single test, set <tt>TESTONE</tt> to its path (relative to
 208 <tt>llvm/test</tt>) and make the <tt>check-one</tt> target:</p>
 209
 210 <div class="doc_code">
 211 <pre>
 212 % gmake TESTONE=Feature/basictest.ll check-one
 213 </pre>
 214 </div>
 215
 216 <p>To run the tests with Valgrind (Memcheck by default), just append
 217 <tt>VG=1</tt> to the commands above, e.g.:</p>
 218
 219 <div class="doc_code">
 220 <pre>
 221 % gmake check VG=1
 222 </pre>
 223 </div>
 224
 225 <!-- _______________________________________________________________________ -->
 226 <div class="doc_subsection"><a name="quicktestsuite">Test suite</a></div>
 227 <!-- _______________________________________________________________________ -->
 228
 229 <p>To run the comprehensive test suite (tests that compile and execute whole
 230 programs), first checkout and setup the <tt>test-suite</tt> module:</p>
 231
 232 <div class="doc_code">
 233 <pre>
 234 % cd llvm/projects
 235 % svn co http://llvm.org/svn/llvm-project/test-suite/trunk test-suite
 236 % cd ..
 237 % ./configure --with-llvmgccdir=$LLVM_GCC_DIR
 238 </pre>
 239 </div>
 240
 241 <p>where <tt>$LLVM_GCC_DIR</tt> is the directory where
 242 you <em>installed</em> llvm-gcc, not it's src or obj
 243 dir. The <tt>--with-llvmgccdir</tt> option assumes that
 244 the <tt>llvm-gcc-4.2</tt> module was configured with
 245 <tt>--program-prefix=llvm-</tt>, and therefore that the C and C++
 246 compiler drivers are called <tt>llvm-gcc</tt> and <tt>llvm-g++</tt>
 247 respectively.  If this is not the case,
 248 use <tt>--with-llvmgcc</tt>/<tt>--with-llvmgxx</tt> to specify each
 249 executable's location.</p>
 250
 251 <p>Then, run the entire test suite by running make in the <tt>test-suite</tt>
 252 directory:</p>
 253
 254 <div class="doc_code">
 255 <pre>
 256 % cd projects/test-suite
 257 % gmake
 258 </pre>
 259 </div>
 260
 261 <p>Usually, running the "nightly" set of tests is a good idea, and you can also
 262 let it generate a report by running:</p>
 263
 264 <div class="doc_code">
 265 <pre>
 266 % cd projects/test-suite
 267 % gmake TEST=nightly report report.html
 268 </pre>
 269 </div>
 270
 271 <p>Any of the above commands can also be run in a subdirectory of
 272 <tt>projects/test-suite</tt> to run the specified test only on the programs in
 273 that subdirectory.</p>
 274
 275 </div>
 276
 277 <!--=========================================================================-->
 278 <div class="doc_section"><a name="dgstructure">DejaGNU structure</a></div>
 279 <!--=========================================================================-->
 280 <div class="doc_text">
 281   <p>The LLVM DejaGNU tests are driven by DejaGNU together with GNU Make and are
 282   located in the <tt>llvm/test</tt> directory.
 283
 284   <p>This directory contains a large array of small tests
 285   that exercise various features of LLVM and to ensure that regressions do not
 286   occur. The directory is broken into several sub-directories, each focused on
 287   a particular area of LLVM. A few of the important ones are:</p>
 288
 289   <ul>
 290     <li><tt>Analysis</tt>: checks Analysis passes.</li>
 291     <li><tt>Archive</tt>: checks the Archive library.</li>
 292     <li><tt>Assembler</tt>: checks Assembly reader/writer functionality.</li>
 293     <li><tt>Bitcode</tt>: checks Bitcode reader/writer functionality.</li>
 294     <li><tt>CodeGen</tt>: checks code generation and each target.</li>
 295     <li><tt>Features</tt>: checks various features of the LLVM language.</li>
 296     <li><tt>Linker</tt>: tests bitcode linking.</li>
 297     <li><tt>Transforms</tt>: tests each of the scalar, IPO, and utility
 298     transforms to ensure they make the right transformations.</li>
 299     <li><tt>Verifier</tt>: tests the IR verifier.</li>
 300   </ul>
 301
 302 </div>
 303
 304 <!-- _______________________________________________________________________ -->
 305 <div class="doc_subsection"><a name="dgcustom">Writing new DejaGNU tests</a></div>
 306 <!-- _______________________________________________________________________ -->
 307 <div class="doc_text">
 308   <p>The DejaGNU structure is very simple, but does require some information to
 309   be set. This information is gathered via <tt>configure</tt> and is written
 310   to a file, <tt>site.exp</tt> in <tt>llvm/test</tt>. The <tt>llvm/test</tt>
 311   Makefile does this work for you.</p>
 312
 313   <p>In order for DejaGNU to work, each directory of tests must have a
 314   <tt>dg.exp</tt> file. DejaGNU looks for this file to determine how to run the
 315   tests. This file is just a Tcl script and it can do anything you want, but
 316   we've standardized it for the LLVM regression tests. If you're adding a
 317   directory of tests, just copy <tt>dg.exp</tt> from another directory to get
 318   running. The standard <tt>dg.exp</tt> simply loads a Tcl
 319   library (<tt>test/lib/llvm.exp</tt>) and calls the <tt>llvm_runtests</tt>
 320   function defined in that library with a list of file names to run. The names
 321   are obtained by using Tcl's glob command.  Any directory that contains only
 322   directories does not need the <tt>dg.exp</tt> file.</p>
 323
 324   <p>The <tt>llvm-runtests</tt> function lookas at each file that is passed to
 325   it and gathers any lines together that match "RUN:". This are the "RUN" lines
 326   that specify how the test is to be run. So, each test script must contain
 327   RUN lines if it is to do anything. If there are no RUN lines, the
 328   <tt>llvm-runtests</tt> function will issue an error and the test will
 329   fail.</p>
 330
 331   <p>RUN lines are specified in the comments of the test program using the
 332   keyword <tt>RUN</tt> followed by a colon, and lastly the command (pipeline)
 333   to execute.  Together, these lines form the "script" that
 334   <tt>llvm-runtests</tt> executes to run the test case.  The syntax of the
 335   RUN lines is similar to a shell's syntax for pipelines including I/O
 336   redirection and variable substitution.  However, even though these lines
 337   may <i>look</i> like a shell script, they are not. RUN lines are interpreted
 338   directly by the Tcl <tt>exec</tt> command. They are never executed by a
 339   shell. Consequently the syntax differs from normal shell script syntax in a
 340   few ways.  You can specify as many RUN lines as needed.</p>
 341
 342   <p>Each RUN line is executed on its own, distinct from other lines unless
 343   its last character is <tt>\</tt>. This continuation character causes the RUN
 344   line to be concatenated with the next one. In this way you can build up long
 345   pipelines of commands without making huge line lengths. The lines ending in
 346   <tt>\</tt> are concatenated until a RUN line that doesn't end in <tt>\</tt> is
 347   found. This concatenated set of RUN lines then constitutes one execution.
 348   Tcl will substitute variables and arrange for the pipeline to be executed. If
 349   any process in the pipeline fails, the entire line (and test case) fails too.
 350   </p>
 351
 352   <p> Below is an example of legal RUN lines in a <tt>.ll</tt> file:</p>
 353
 354 <div class="doc_code">
 355 <pre>
 356 ; RUN: llvm-as &lt; %s | llvm-dis &gt; %t1
 357 ; RUN: llvm-dis &lt; %s.bc-13 &gt; %t2
 358 ; RUN: diff %t1 %t2
 359 </pre>
 360 </div>
 361
 362   <p>As with a Unix shell, the RUN: lines permit pipelines and I/O redirection
 363   to be used. However, the usage is slightly different than for Bash. To check
 364   what's legal, see the documentation for the
 365   <a href="http://www.tcl.tk/man/tcl8.5/TclCmd/exec.htm#M2">Tcl exec</a>
 366   command and the
 367   <a href="http://www.tcl.tk/man/tcl8.5/tutorial/Tcl26.html">tutorial</a>.
 368   The major differences are:</p>
 369   <ul>
 370     <li>You can't do <tt>2&gt;&amp;1</tt>. That will cause Tcl to write to a
 371     file named <tt>&amp;1</tt>. Usually this is done to get stderr to go through
 372     a pipe. You can do that in tcl with <tt>|&amp;</tt> so replace this idiom:
 373     <tt>... 2&gt;&amp;1 | grep</tt> with <tt>... |&amp; grep</tt></li>
 374     <li>You can only redirect to a file, not to another descriptor and not from
 375     a here document.</li>
 376     <li>tcl supports redirecting to open files with the @ syntax but you
 377     shouldn't use that here.</li>
 378   </ul>
 379
 380   <p>There are some quoting rules that you must pay attention to when writing
 381   your RUN lines. In general nothing needs to be quoted. Tcl won't strip off any
 382   ' or " so they will get passed to the invoked program. For example:</p>
 383
 384 <div class="doc_code">
 385 <pre>
 386 ... | grep 'find this string'
 387 </pre>
 388 </div>
 389
 390   <p>This will fail because the ' characters are passed to grep. This would
 391   instruction grep to look for <tt>'find</tt> in the files <tt>this</tt> and
 392   <tt>string'</tt>. To avoid this use curly braces to tell Tcl that it should
 393   treat everything enclosed as one value. So our example would become:</p>
 394
 395 <div class="doc_code">
 396 <pre>
 397 ... | grep {find this string}
 398 </pre>
 399 </div>
 400
 401   <p>Additionally, the characters <tt>[</tt> and <tt>]</tt> are treated
 402   specially by Tcl. They tell Tcl to interpret the content as a command to
 403   execute. Since these characters are often used in regular expressions this can
 404   have disastrous results and cause the entire test run in a directory to fail.
 405   For example, a common idiom is to look for some basicblock number:</p>
 406
 407 <div class="doc_code">
 408 <pre>
 409 ... | grep bb[2-8]
 410 </pre>
 411 </div>
 412
 413   <p>This, however, will cause Tcl to fail because its going to try to execute
 414   a program named "2-8". Instead, what you want is this:</p>
 415
 416 <div class="doc_code">
 417 <pre>
 418 ... | grep {bb\[2-8\]}
 419 </pre>
 420 </div>
 421
 422   <p>Finally, if you need to pass the <tt>\</tt> character down to a program,
 423   then it must be doubled. This is another Tcl special character. So, suppose
 424   you had:
 425
 426 <div class="doc_code">
 427 <pre>
 428 ... | grep 'i32\*'
 429 </pre>
 430 </div>
 431
 432   <p>This will fail to match what you want (a pointer to i32). First, the
 433   <tt>'</tt> do not get stripped off. Second, the <tt>\</tt> gets stripped off
 434   by Tcl so what grep sees is: <tt>'i32*'</tt>. That's not likely to match
 435   anything. To resolve this you must use <tt>\\</tt> and the <tt>{}</tt>, like
 436   this:</p>
 437
 438 <div class="doc_code">
 439 <pre>
 440 ... | grep {i32\\*}
 441 </pre>
 442 </div>
 443
 444 <p>If your system includes GNU <tt>grep</tt>, make sure
 445 that <tt>GREP_OPTIONS</tt> is not set in your environment. Otherwise,
 446 you may get invalid results (both false positives and false
 447 negatives).</p>
 448
 449 </div>
 450
 451 <!-- _______________________________________________________________________ -->
 452 <div class="doc_subsection"><a name="FileCheck">The FileCheck utility</a></div>
 453 <!-- _______________________________________________________________________ -->
 454
 455 <div class="doc_text">
 456
 457 <p>A powerful feature of the RUN: lines is that it allows any arbitrary commands
 458    to be executed as part of the test harness.  While standard (portable) unix
 459    tools like 'grep' work fine on run lines, as you see above, there are a lot
 460    of caveats due to interaction with Tcl syntax, and we want to make sure the
 461    run lines are portable to a wide range of systems.  Another major problem is
 462    that grep is not very good at checking to verify that the output of a tools
 463    contains a series of different output in a specific order.  The FileCheck
 464    tool was designed to help with these problems.</p>
 465
 466 <p>FileCheck (whose basic command line arguments are described in <a
 467    href="http://llvm.org/cmds/FileCheck.html">the FileCheck man page</a> is
 468    designed to read a file to check from standard input, and the set of things
 469    to verify from a file specified as a command line argument.  A simple example
 470    of using FileCheck from a RUN line looks like this:</p>
 471
 472 <div class="doc_code">
 473 <pre>
 474 ; RUN: llvm-as &lt; %s | llc -march=x86-64 | <b>FileCheck %s</b>
 475 </pre>
 476 </div>
 477
 478 <p>This syntax says to pipe the current file ("%s") into llvm-as, pipe that into
 479 llc, then pipe the output of llc into FileCheck.  This means that FileCheck will
 480 be verifying its standard input (the llc output) against the filename argument
 481 specified (the original .ll file specified by "%s").  To see how this works,
 482 lets look at the rest of the .ll file (after the RUN line):</p>
 483
 484 <div class="doc_code">
 485 <pre>
 486 define void @sub1(i32* %p, i32 %v) {
 487 entry:
 488 ; <b>CHECK: sub1:</b>
 489 ; <b>CHECK: subl</b>
 490         %0 = tail call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %p, i32 %v)
 491         ret void
 492 }
 493
 494 define void @inc4(i64* %p) {
 495 entry:
 496 ; <b>CHECK: inc4:</b>
 497 ; <b>CHECK: incq</b>
 498         %0 = tail call i64 @llvm.atomic.load.add.i64.p0i64(i64* %p, i64 1)
 499         ret void
 500 }
 501 </pre>
 502 </div>
 503
 504 <p>Here you can see some "CHECK:" lines specified in comments.  Now you can see
 505 how the file is piped into llvm-as, then llc, and the machine code output is
 506 what we are verifying.  FileCheck checks the machine code output to verify that
 507 it matches what the "CHECK:" lines specify.</p>
 508
 509 <p>The syntax of the CHECK: lines is very simple: they are fixed strings that
 510 must occur in order.  FileCheck defaults to ignoring horizontal whitespace
 511 differences (e.g. a space is allowed to match a tab) but otherwise, the contents
 512 of the CHECK: line is required to match some thing in the test file exactly.</p>
 513
 514 <p>One nice thing about FileCheck (compared to grep) is that it allows merging
 515 test cases together into logical groups.  For example, because the test above
 516 is checking for the "sub1:" and "inc4:" labels, it will not match unless there
 517 is a "subl" in between those labels.  If it existed somewhere else in the file,
 518 that would not count: "grep subl" matches if subl exists anywhere in the
 519 file.</p>
 520
 521 </div>
 522
 523 <!-- _______________________________________________________________________ -->
 524 <div class="doc_subsubsection"><a
 525 name="FileCheck-check-prefix">The FileCheck -check-prefix option</a></div>
 526
 527 <div class="doc_text">
 528
 529 <p>The FileCheck -check-prefix option allows multiple test configurations to be
 530 driven from one .ll file.  This is useful in many circumstances, for example,
 531 testing different architectural variants with llc.  Here's a simple example:</p>
 532
 533 <div class="doc_code">
 534 <pre>
 535 ; RUN: llvm-as &lt; %s | llc -mtriple=i686-apple-darwin9 -mattr=sse41 \
 536 ; RUN:              | <b>FileCheck %s -check-prefix=X32</b>
 537 ; RUN: llvm-as &lt; %s | llc -mtriple=x86_64-apple-darwin9 -mattr=sse41 \
 538 ; RUN:              | <b>FileCheck %s -check-prefix=X64</b>
 539
 540 define &lt;4 x i32&gt; @pinsrd_1(i32 %s, &lt;4 x i32&gt; %tmp) nounwind {
 541         %tmp1 = insertelement &lt;4 x i32&gt; %tmp, i32 %s, i32 1
 542         ret &lt;4 x i32&gt; %tmp1
 543 ; <b>X32:</b> pinsrd_1:
 544 ; <b>X32:</b>    pinsrd $1, 4(%esp), %xmm0
 545
 546 ; <b>X64:</b> pinsrd_1:
 547 ; <b>X64:</b>    pinsrd $1, %edi, %xmm0
 548 }
 549 </pre>
 550 </div>
 551
 552 <p>In this case, we're testing that we get the expected code generation with
 553 both 32-bit and 64-bit code generation.</p>
 554
 555 </div>
 556
 557 <!-- _______________________________________________________________________ -->
 558 <div class="doc_subsubsection"><a
 559 name="FileCheck-CHECK-NEXT">The "CHECK-NEXT:" directive</a></div>
 560
 561 <div class="doc_text">
 562
 563 <p>Sometimes you want to match lines and would like to verify that matches
 564 happen on exactly consequtive lines with no other lines in between them.  In
 565 this case, you can use CHECK: and CHECK-NEXT: directives to specify this.  If
 566 you specified a custom check prefix, just use "&lt;PREFIX&gt;-NEXT:".  For
 567 example, something like this works as you'd expect:</p>
 568
 569 <div class="doc_code">
 570 <pre>
 571 define void @t2(&lt;2 x double&gt;* %r, &lt;2 x double&gt;* %A, double %B) {
 572         %tmp3 = load &lt;2 x double&gt;* %A, align 16
 573         %tmp7 = insertelement &lt;2 x double&gt; undef, double %B, i32 0
 574         %tmp9 = shufflevector &lt;2 x double&gt; %tmp3,
 575                               &lt;2 x double&gt; %tmp7,
 576                               &lt;2 x i32&gt; &lt; i32 0, i32 2 &gt;
 577         store &lt;2 x double&gt; %tmp9, &lt;2 x double&gt;* %r, align 16
 578         ret void
 579
 580 ; <b>CHECK:</b> t2:
 581 ; <b>CHECK:</b>         movl    8(%esp), %eax
 582 ; <b>CHECK-NEXT:</b>    movapd  (%eax), %xmm0
 583 ; <b>CHECK-NEXT:</b>    movhpd  12(%esp), %xmm0
 584 ; <b>CHECK-NEXT:</b>    movl    4(%esp), %eax
 585 ; <b>CHECK-NEXT:</b>    movapd  %xmm0, (%eax)
 586 ; <b>CHECK-NEXT:</b>    ret
 587 }
 588 </pre>
 589 </div>
 590
 591 <p>CHECK-NEXT: directives reject the input unless there is exactly one newline
 592 between it an the previous directive.  A CHECK-NEXT cannot be the first
 593 directive in a file.</p>
 594
 595 </div>
 596
 597 <!-- _______________________________________________________________________ -->
 598 <div class="doc_subsection"><a name="dgvars">Variables and
 599 substitutions</a></div>
 600 <!-- _______________________________________________________________________ -->
 601 <div class="doc_text">
 602   <p>With a RUN line there are a number of substitutions that are permitted. In
 603   general, any Tcl variable that is available in the <tt>substitute</tt>
 604   function (in <tt>test/lib/llvm.exp</tt>) can be substituted into a RUN line.
 605   To make a substitution just write the variable's name preceded by a $.
 606   Additionally, for compatibility reasons with previous versions of the test
 607   library, certain names can be accessed with an alternate syntax: a % prefix.
 608   These alternates are deprecated and may go away in a future version.
 609   </p>
 610   <p>Here are the available variable names. The alternate syntax is listed in
 611   parentheses.</p>
 612
 613   <dl style="margin-left: 25px">
 614     <dt><b>$test</b> (%s)</dt>
 615     <dd>The full path to the test case's source. This is suitable for passing
 616     on the command line as the input to an llvm tool.</dd>
 617
 618     <dt><b>$srcdir</b></dt>
 619     <dd>The source directory from where the "<tt>make check</tt>" was run.</dd>
 620
 621     <dt><b>objdir</b></dt>
 622     <dd>The object directory that corresponds to the <tt>$srcdir</tt>.</dd>
 623
 624     <dt><b>subdir</b></dt>
 625     <dd>A partial path from the <tt>test</tt> directory that contains the
 626     sub-directory that contains the test source being executed.</dd>
 627
 628     <dt><b>srcroot</b></dt>
 629     <dd>The root directory of the LLVM src tree.</dd>
 630
 631     <dt><b>objroot</b></dt>
 632     <dd>The root directory of the LLVM object tree. This could be the same
 633     as the srcroot.</dd>
 634
 635     <dt><b>path</b><dt>
 636     <dd>The path to the directory that contains the test case source.  This is
 637     for locating any supporting files that are not generated by the test, but
 638     used by the test.</dd>
 639
 640     <dt><b>tmp</b></dt>
 641     <dd>The path to a temporary file name that could be used for this test case.
 642     The file name won't conflict with other test cases. You can append to it if
 643     you need multiple temporaries. This is useful as the destination of some
 644     redirected output.</dd>
 645
 646     <dt><b>llvmlibsdir</b> (%llvmlibsdir)</dt>
 647     <dd>The directory where the LLVM libraries are located.</dd>
 648
 649     <dt><b>target_triplet</b> (%target_triplet)</dt>
 650     <dd>The target triplet that corresponds to the current host machine (the one
 651     running the test cases). This should probably be called "host".<dd>
 652
 653     <dt><b>llvmgcc</b> (%llvmgcc)</dt>
 654     <dd>The full path to the <tt>llvm-gcc</tt> executable as specified in the
 655     configured LLVM environment</dd>
 656
 657     <dt><b>llvmgxx</b> (%llvmgxx)</dt>
 658     <dd>The full path to the <tt>llvm-gxx</tt> executable as specified in the
 659     configured LLVM environment</dd>
 660
 661     <dt><b>llvmgcc_version</b> (%llvmgcc_version)</dt>
 662     <dd>The full version number of the <tt>llvm-gcc</tt> executable.</dd>
 663
 664     <dt><b>llvmgccmajvers</b> (%llvmgccmajvers)</dt>
 665     <dd>The major version number of the <tt>llvm-gcc</tt> executable.</dd>
 666
 667     <dt><b>gccpath</b></dt>
 668     <dd>The full path to the C compiler used to <i>build </i> LLVM. Note that
 669     this might not be gcc.</dd>
 670
 671     <dt><b>gxxpath</b></dt>
 672     <dd>The full path to the C++ compiler used to <i>build </i> LLVM. Note that
 673     this might not be g++.</dd>
 674
 675     <dt><b>compile_c</b> (%compile_c)</dt>
 676     <dd>The full command line used to compile LLVM C source  code. This has all
 677     the configured -I, -D and optimization options.</dd>
 678
 679     <dt><b>compile_cxx</b> (%compile_cxx)</dt>
 680     <dd>The full command used to compile LLVM C++ source  code. This has
 681     all the configured -I, -D and optimization options.</dd>
 682
 683     <dt><b>link</b> (%link)</dt>
 684     <dd>This full link command used to link LLVM executables. This has all the
 685     configured -I, -L and -l options.</dd>
 686
 687     <dt><b>shlibext</b> (%shlibext)</dt>
 688     <dd>The suffix for the host platforms share library (dll) files. This
 689     includes the period as the first character.</dd>
 690   </dl>
 691   <p>To add more variables, two things need to be changed. First, add a line in
 692   the <tt>test/Makefile</tt> that creates the <tt>site.exp</tt> file. This will
 693   "set" the variable as a global in the site.exp file. Second, in the
 694   <tt>test/lib/llvm.exp</tt> file, in the substitute proc, add the variable name
 695   to the list of "global" declarations at the beginning of the proc. That's it,
 696   the variable can then be used in test scripts.</p>
 697 </div>
 698
 699 <!-- _______________________________________________________________________ -->
 700 <div class="doc_subsection"><a name="dgfeatures">Other Features</a></div>
 701 <!-- _______________________________________________________________________ -->
 702 <div class="doc_text">
 703   <p>To make RUN line writing easier, there are several shell scripts located
 704   in the <tt>llvm/test/Scripts</tt> directory. This directory is in the PATH
 705   when running tests, so you can just call these scripts using their name. For
 706   example:</p>
 707   <dl>
 708     <dt><b>ignore</b></dt>
 709     <dd>This script runs its arguments and then always returns 0. This is useful
 710     in cases where the test needs to cause a tool to generate an error (e.g. to
 711     check the error output). However, any program in a pipeline that returns a
 712     non-zero result will cause the test to fail. This script overcomes that
 713     issue and nicely documents that the test case is purposefully ignoring the
 714     result code of the tool</dd>
 715
 716     <dt><b>not</b></dt>
 717     <dd>This script runs its arguments and then inverts the result code from
 718     it. Zero result codes become 1. Non-zero result codes become 0. This is
 719     useful to invert the result of a grep. For example "not grep X" means
 720     succeed only if you don't find X in the input.</dd>
 721   </dl>
 722
 723   <p>Sometimes it is necessary to mark a test case as "expected fail" or XFAIL.
 724   You can easily mark a test as XFAIL just by including  <tt>XFAIL: </tt> on a
 725   line near the top of the file. This signals that the test case should succeed
 726   if the test fails. Such test cases are counted separately by DejaGnu. To
 727   specify an expected fail, use the XFAIL keyword in the comments of the test
 728   program followed by a colon and one or more regular expressions (separated by
 729   a comma). The regular expressions allow you to XFAIL the test conditionally
 730   by host platform. The regular expressions following the : are matched against
 731   the target triplet or llvmgcc version number for the host machine. If there is
 732   a match, the test is expected to fail. If not, the test is expected to
 733   succeed. To XFAIL everywhere just specify <tt>XFAIL: *</tt>. When matching
 734   the llvm-gcc version, you can specify the major (e.g. 3) or full version
 735   (i.e. 3.4) number. Here is an example of an <tt>XFAIL</tt> line:</p>
 736
 737 <div class="doc_code">
 738 <pre>
 739 ; XFAIL: darwin,sun,llvmgcc4
 740 </pre>
 741 </div>
 742
 743   <p>To make the output more useful, the <tt>llvm_runtest</tt> function wil
 744   scan the lines of the test case for ones that contain a pattern that matches
 745   PR[0-9]+. This is the syntax for specifying a PR (Problem Report) number that
 746   is related to the test case. The number after "PR" specifies the LLVM bugzilla
 747   number. When a PR number is specified, it will be used in the pass/fail
 748   reporting. This is useful to quickly get some context when a test fails.</p>
 749
 750   <p>Finally, any line that contains "END." will cause the special
 751   interpretation of lines to terminate. This is generally done right after the
 752   last RUN: line. This has two side effects: (a) it prevents special
 753   interpretation of lines that are part of the test program, not the
 754   instructions to the test case, and (b) it speeds things up for really big test
 755   cases by avoiding interpretation of the remainder of the file.</p>
 756
 757 </div>
 758
 759 <!--=========================================================================-->
 760 <div class="doc_section"><a name="testsuitestructure">Test suite
 761 Structure</a></div>
 762 <!--=========================================================================-->
 763
 764 <div class="doc_text">
 765
 766 <p>The <tt>test-suite</tt> module contains a number of programs that can be compiled
 767 with LLVM and executed. These programs are compiled using the native compiler
 768 and various LLVM backends. The output from the program compiled with the
 769 native compiler is assumed correct; the results from the other programs are
 770 compared to the native program output and pass if they match.</p>
 771
 772 <p>When executing tests, it is usually a good idea to start out with a subset of
 773 the available tests or programs. This makes test run times smaller at first and
 774 later on this is useful to investigate individual test failures. To run some
 775 test only on a subset of programs, simply change directory to the programs you
 776 want tested and run <tt>gmake</tt> there. Alternatively, you can run a different
 777 test using the <tt>TEST</tt> variable to change what tests or run on the
 778 selected programs (see below for more info).</p>
 779
 780 <p>In addition for testing correctness, the <tt>llvm-test</tt> directory also
 781 performs timing tests of various LLVM optimizations.  It also records
 782 compilation times for the compilers and the JIT.  This information can be
 783 used to compare the effectiveness of LLVM's optimizations and code
 784 generation.</p>
 785
 786 <p><tt>llvm-test</tt> tests are divided into three types of tests: MultiSource,
 787 SingleSource, and External.</p>
 788
 789 <ul>
 790 <li><tt>llvm-test/SingleSource</tt>
 791 <p>The SingleSource directory contains test programs that are only a single
 792 source file in size.  These are usually small benchmark programs or small
 793 programs that calculate a particular value.  Several such programs are grouped
 794 together in each directory.</p></li>
 795
 796 <li><tt>llvm-test/MultiSource</tt>
 797 <p>The MultiSource directory contains subdirectories which contain entire
 798 programs with multiple source files.  Large benchmarks and whole applications
 799 go here.</p></li>
 800
 801 <li><tt>llvm-test/External</tt>
 802 <p>The External directory contains Makefiles for building code that is external
 803 to (i.e., not distributed with) LLVM.  The most prominent members of this
 804 directory are the SPEC 95 and SPEC 2000 benchmark suites. The <tt>External</tt>
 805 directory does not contain these actual tests, but only the Makefiles that know
 806 how to properly compile these programs from somewhere else. The presence and
 807 location of these external programs is configured by the llvm-test
 808 <tt>configure</tt> script.</p></li>
 809 </ul>
 810
 811 <p>Each tree is then subdivided into several categories, including applications,
 812 benchmarks, regression tests, code that is strange grammatically, etc.  These
 813 organizations should be relatively self explanatory.</p>
 814
 815 <p>Some tests are known to fail.  Some are bugs that we have not fixed yet;
 816 others are features that we haven't added yet (or may never add).  In DejaGNU,
 817 the result for such tests will be XFAIL (eXpected FAILure).  In this way, you
 818 can tell the difference between an expected and unexpected failure.</p>
 819
 820 <p>The tests in the test suite have no such feature at this time. If the
 821 test passes, only warnings and other miscellaneous output will be generated.  If
 822 a test fails, a large &lt;program&gt; FAILED message will be displayed.  This
 823 will help you separate benign warnings from actual test failures.</p>
 824
 825 </div>
 826
 827 <!--=========================================================================-->
 828 <div class="doc_section"><a name="testsuiterun">Running the test suite</a></div>
 829 <!--=========================================================================-->
 830
 831 <div class="doc_text">
 832
 833 <p>First, all tests are executed within the LLVM object directory tree.  They
 834 <i>are not</i> executed inside of the LLVM source tree. This is because the
 835 test suite creates temporary files during execution.</p>
 836
 837 <p>To run the test suite, you need to use the following steps:</p>
 838
 839 <ol>
 840   <li><tt>cd</tt> into the <tt>llvm/projects</tt> directory in your source tree.
 841   </li>
 842
 843   <li><p>Check out the <tt>test-suite</tt> module with:</p>
 844
 845 <div class="doc_code">
 846 <pre>
 847 % svn co http://llvm.org/svn/llvm-project/test-suite/trunk test-suite
 848 </pre>
 849 </div>
 850     <p>This will get the test suite into <tt>llvm/projects/test-suite</tt>.</p>
 851   </li>
 852   <li><p>Configure and build <tt>llvm</tt>.</p></li>
 853   <li><p>Configure and build <tt>llvm-gcc</tt>.</p></li>
 854   <li><p>Install <tt>llvm-gcc</tt> somewhere.</p></li>
 855   <li><p><em>Re-configure</em> <tt>llvm</tt> from the top level of
 856       each build tree (LLVM object directory tree) in which you want
 857       to run the test suite, just as you do before building LLVM.</p>
 858     <p>During the <em>re-configuration</em>, you must either: (1)
 859       have <tt>llvm-gcc</tt> you just built in your path, or (2)
 860       specify the directory where your just-built <tt>llvm-gcc</tt> is
 861       installed using <tt>--with-llvmgccdir=$LLVM_GCC_DIR</tt>.</p>
 862     <p>You must also tell the configure machinery that the test suite
 863       is available so it can be configured for your build tree:</p>
 864 <div class="doc_code">
 865 <pre>
 866 % cd $LLVM_OBJ_ROOT ; $LLVM_SRC_ROOT/configure [--with-llvmgccdir=$LLVM_GCC_DIR]
 867 </pre>
 868 </div>
 869     <p>[Remember that <tt>$LLVM_GCC_DIR</tt> is the directory where you
 870     <em>installed</em> llvm-gcc, not its src or obj directory.]</p>
 871   </li>
 872
 873   <li><p>You can now run the test suite from your build tree as follows:</p>
 874 <div class="doc_code">
 875 <pre>
 876 % cd $LLVM_OBJ_ROOT/projects/test-suite
 877 % make
 878 </pre>
 879 </div>
 880   </li>
 881 </ol>
 882 <p>Note that the second and third steps only need to be done once. After you
 883 have the suite checked out and configured, you don't need to do it again (unless
 884 the test code or configure script changes).</p>
 885
 886 </div>
 887
 888 <!-- _______________________________________________________________________ -->
 889 <div class="doc_subsection">
 890 <a name="testsuiteexternal">Configuring External Tests</a></div>
 891 <!-- _______________________________________________________________________ -->
 892
 893 <div class="doc_text">
 894 <p>In order to run the External tests in the <tt>test-suite</tt>
 895   module, you must specify <i>--with-externals</i>.  This
 896   must be done during the <em>re-configuration</em> step (see above),
 897   and the <tt>llvm</tt> re-configuration must recognize the
 898   previously-built <tt>llvm-gcc</tt>.  If any of these is missing or
 899   neglected, the External tests won't work.</p>
 900 <dl>
 901 <dt><i>--with-externals</i></dt>
 902 <dt><i>--with-externals=&lt;<tt>directory</tt>&gt;</i></dt>
 903 </dl>
 904   This tells LLVM where to find any external tests.  They are expected to be
 905   in specifically named subdirectories of &lt;<tt>directory</tt>&gt;.
 906   If <tt>directory</tt> is left unspecified,
 907   <tt>configure</tt> uses the default value
 908   <tt>/home/vadve/shared/benchmarks/speccpu2000/benchspec</tt>.
 909   Subdirectory names known to LLVM include:
 910   <dl>
 911   <dt>spec95</dt>
 912   <dt>speccpu2000</dt>
 913   <dt>speccpu2006</dt>
 914   <dt>povray31</dt>
 915   </dl>
 916   Others are added from time to time, and can be determined from
 917   <tt>configure</tt>.
 918 </div>
 919
 920 <!-- _______________________________________________________________________ -->
 921 <div class="doc_subsection">
 922 <a name="testsuitetests">Running different tests</a></div>
 923 <!-- _______________________________________________________________________ -->
 924 <div class="doc_text">
 925 <p>In addition to the regular "whole program" tests, the <tt>test-suite</tt>
 926 module also provides a mechanism for compiling the programs in different ways.
 927 If the variable TEST is defined on the <tt>gmake</tt> command line, the test system will
 928 include a Makefile named <tt>TEST.&lt;value of TEST variable&gt;.Makefile</tt>.
 929 This Makefile can modify build rules to yield different results.</p>
 930
 931 <p>For example, the LLVM nightly tester uses <tt>TEST.nightly.Makefile</tt> to
 932 create the nightly test reports.  To run the nightly tests, run <tt>gmake
 933 TEST=nightly</tt>.</p>
 934
 935 <p>There are several TEST Makefiles available in the tree.  Some of them are
 936 designed for internal LLVM research and will not work outside of the LLVM
 937 research group.  They may still be valuable, however, as a guide to writing your
 938 own TEST Makefile for any optimization or analysis passes that you develop with
 939 LLVM.</p>
 940
 941 </div>
 942
 943 <!-- _______________________________________________________________________ -->
 944 <div class="doc_subsection">
 945 <a name="testsuiteoutput">Generating test output</a></div>
 946 <!-- _______________________________________________________________________ -->
 947 <div class="doc_text">
 948   <p>There are a number of ways to run the tests and generate output. The most
 949   simple one is simply running <tt>gmake</tt> with no arguments. This will
 950   compile and run all programs in the tree using a number of different methods
 951   and compare results. Any failures are reported in the output, but are likely
 952   drowned in the other output. Passes are not reported explicitely.</p>
 953
 954   <p>Somewhat better is running <tt>gmake TEST=sometest test</tt>, which runs
 955   the specified test and usually adds per-program summaries to the output
 956   (depending on which sometest you use). For example, the <tt>nightly</tt> test
 957   explicitely outputs TEST-PASS or TEST-FAIL for every test after each program.
 958   Though these lines are still drowned in the output, it's easy to grep the
 959   output logs in the Output directories.</p>
 960
 961   <p>Even better are the <tt>report</tt> and <tt>report.format</tt> targets
 962   (where <tt>format</tt> is one of <tt>html</tt>, <tt>csv</tt>, <tt>text</tt> or
 963   <tt>graphs</tt>). The exact contents of the report are dependent on which
 964   <tt>TEST</tt> you are running, but the text results are always shown at the
 965   end of the run and the results are always stored in the
 966   <tt>report.&lt;type&gt;.format</tt> file (when running with
 967   <tt>TEST=&lt;type&gt;</tt>).
 968
 969   The <tt>report</tt> also generate a file called
 970   <tt>report.&lt;type&gt;.raw.out</tt> containing the output of the entire test
 971   run.
 972 </div>
 973
 974 <!-- _______________________________________________________________________ -->
 975 <div class="doc_subsection">
 976 <a name="testsuitecustom">Writing custom tests for the test suite</a></div>
 977 <!-- _______________________________________________________________________ -->
 978
 979 <div class="doc_text">
 980
 981 <p>Assuming you can run the test suite, (e.g. "<tt>gmake TEST=nightly report</tt>"
 982 should work), it is really easy to run optimizations or code generator
 983 components against every program in the tree, collecting statistics or running
 984 custom checks for correctness.  At base, this is how the nightly tester works,
 985 it's just one example of a general framework.</p>
 986
 987 <p>Lets say that you have an LLVM optimization pass, and you want to see how
 988 many times it triggers.  First thing you should do is add an LLVM
 989 <a href="ProgrammersManual.html#Statistic">statistic</a> to your pass, which
 990 will tally counts of things you care about.</p>
 991
 992 <p>Following this, you can set up a test and a report that collects these and
 993 formats them for easy viewing.  This consists of two files, an
 994 "<tt>test-suite/TEST.XXX.Makefile</tt>" fragment (where XXX is the name of your
 995 test) and an "<tt>llvm-test/TEST.XXX.report</tt>" file that indicates how to
 996 format the output into a table.  There are many example reports of various
 997 levels of sophistication included with the test suite, and the framework is very
 998 general.</p>
 999
1000 <p>If you are interested in testing an optimization pass, check out the
1001 "libcalls" test as an example.  It can be run like this:<p>
1002
1003 <div class="doc_code">
1004 <pre>
1005 % cd llvm/projects/test-suite/MultiSource/Benchmarks  # or some other level
1006 % make TEST=libcalls report
1007 </pre>
1008 </div>
1009
1010 <p>This will do a bunch of stuff, then eventually print a table like this:</p>
1011
1012 <div class="doc_code">
1013 <pre>
1014 Name                                  | total | #exit |
1015 ...
1016 FreeBench/analyzer/analyzer           | 51    | 6     |
1017 FreeBench/fourinarow/fourinarow       | 1     | 1     |
1018 FreeBench/neural/neural               | 19    | 9     |
1019 FreeBench/pifft/pifft                 | 5     | 3     |
1020 MallocBench/cfrac/cfrac               | 1     | *     |
1021 MallocBench/espresso/espresso         | 52    | 12    |
1022 MallocBench/gs/gs                     | 4     | *     |
1023 Prolangs-C/TimberWolfMC/timberwolfmc  | 302   | *     |
1024 Prolangs-C/agrep/agrep                | 33    | 12    |
1025 Prolangs-C/allroots/allroots          | *     | *     |
1026 Prolangs-C/assembler/assembler        | 47    | *     |
1027 Prolangs-C/bison/mybison              | 74    | *     |
1028 ...
1029 </pre>
1030 </div>
1031
1032 <p>This basically is grepping the -stats output and displaying it in a table.
1033 You can also use the "TEST=libcalls report.html" target to get the table in HTML
1034 form, similarly for report.csv and report.tex.</p>
1035
1036 <p>The source for this is in test-suite/TEST.libcalls.*.  The format is pretty
1037 simple: the Makefile indicates how to run the test (in this case,
1038 "<tt>opt -simplify-libcalls -stats</tt>"), and the report contains one line for
1039 each column of the output.  The first value is the header for the column and the
1040 second is the regex to grep the output of the command for.  There are lots of
1041 example reports that can do fancy stuff.</p>
1042
1043 </div>
1044
1045
1046 <!--=========================================================================-->
1047 <div class="doc_section"><a name="nightly">Running the nightly tester</a></div>
1048 <!--=========================================================================-->
1049
1050 <div class="doc_text">
1051
1052 <p>
1053 The <a href="http://llvm.org/nightlytest/">LLVM Nightly Testers</a>
1054 automatically check out an LLVM tree, build it, run the "nightly"
1055 program test (described above), run all of the DejaGNU tests,
1056 delete the checked out tree, and then submit the results to
1057 <a href="http://llvm.org/nightlytest/">http://llvm.org/nightlytest/</a>.
1058 After test results are submitted to
1059 <a href="http://llvm.org/nightlytest/">http://llvm.org/nightlytest/</a>,
1060 they are processed and displayed on the tests page. An email to
1061 <a href="http://lists.cs.uiuc.edu/pipermail/llvm-testresults/">
1062 llvm-testresults@cs.uiuc.edu</a> summarizing the results is also generated.
1063 This testing scheme is designed to ensure that programs don't break as well
1064 as keep track of LLVM's progress over time.</p>
1065
1066 <p>If you'd like to set up an instance of the nightly tester to run on your
1067 machine, take a look at the comments at the top of the
1068 <tt>utils/NewNightlyTest.pl</tt> file. If you decide to set up a nightly tester
1069 please choose a unique nickname and invoke <tt>utils/NewNightlyTest.pl</tt>
1070 with the "-nickname [yournickname]" command line option.
1071
1072 <p>You can create a shell script to encapsulate the running of the script.
1073 The optimized x86 Linux nightly test is run from just such a script:</p>
1074
1075 <div class="doc_code">
1076 <pre>
1077 #!/bin/bash
1078 BASE=/proj/work/llvm/nightlytest
1079 export BUILDDIR=$BASE/build
1080 export WEBDIR=$BASE/testresults
1081 export LLVMGCCDIR=/proj/work/llvm/cfrontend/install
1082 export PATH=/proj/install/bin:$LLVMGCCDIR/bin:$PATH
1083 export LD_LIBRARY_PATH=/proj/install/lib
1084 cd $BASE
1085 cp /proj/work/llvm/llvm/utils/NewNightlyTest.pl .
1086 nice ./NewNightlyTest.pl -nice -release -verbose -parallel -enable-linscan \
1087    -nickname NightlyTester -noexternals &gt; output.log 2&gt;&amp;1
1088 </pre>
1089 </div>
1090
1091 <p>It is also possible to specify the the location your nightly test results
1092 are submitted. You can do this by passing the command line option
1093 "-submit-server [server_address]" and "-submit-script [script_on_server]" to
1094 <tt>utils/NewNightlyTest.pl</tt>. For example, to submit to the llvm.org
1095 nightly test results page, you would invoke the nightly test script with
1096 "-submit-server llvm.org -submit-script /nightlytest/NightlyTestAccept.cgi".
1097 If these options are not specified, the nightly test script sends the results
1098 to the llvm.org nightly test results page.</p>
1099
1100 <p>Take a look at the <tt>NewNightlyTest.pl</tt> file to see what all of the
1101 flags and strings do.  If you start running the nightly tests, please let us
1102 know. Thanks!</p>
1103
1104 </div>
1105
1106 <!-- *********************************************************************** -->
1107
1108 <hr>
1109 <address>
1110   <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
1111   src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
1112   <a href="http://validator.w3.org/check/referer"><img
1113   src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
1114
1115   John T. Criswell, Reid Spencer, and Tanya Lattner<br>
1116   <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
1117   Last modified: $Date$
1118 </address>
1119 </body>
1120 </html>