llvm/docs/Bugpoint.html
Joel Jones a9498a2ee6 Fix broken internal link.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@155213 91177308-0d34-0410-b5e6-96231b3b80d8
2012-04-20 18:20:24 +00:00

317 lines
14 KiB
HTML

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<title>LLVM bugpoint tool: design and usage</title>
<link rel="stylesheet" href="_static/llvm.css" type="text/css">
</head>
<h1>
LLVM bugpoint tool: design and usage
</h1>
<ul>
<li><a href="#desc">Description</a></li>
<li><a href="#design">Design Philosophy</a>
<ul>
<li><a href="#autoselect">Automatic Debugger Selection</a></li>
<li><a href="#crashdebug">Crash debugger</a></li>
<li><a href="#codegendebug">Code generator debugger</a></li>
<li><a href="#miscompilationdebug">Miscompilation debugger</a></li>
</ul></li>
<li><a href="#advice">Advice for using <tt>bugpoint</tt></a></li>
<li><a href="#notEnough">What to do when <tt>bugpoint</tt> isn't enough</a></li>
</ul>
<div class="doc_author">
<p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a></p>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="desc">Description</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p><tt>bugpoint</tt> narrows down the source of problems in LLVM tools and
passes. It can be used to debug three types of failures: optimizer crashes,
miscompilations by optimizers, or bad native code generation (including problems
in the static and JIT compilers). It aims to reduce large test cases to small,
useful ones. For example, if <tt>opt</tt> crashes while optimizing a
file, it will identify the optimization (or combination of optimizations) that
causes the crash, and reduce the file down to a small example which triggers the
crash.</p>
<p>For detailed case scenarios, such as debugging <tt>opt</tt>, or one of the
LLVM code generators, see <a href="HowToSubmitABug.html">How To Submit a Bug
Report document</a>.</p>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="design">Design Philosophy</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p><tt>bugpoint</tt> is designed to be a useful tool without requiring any
hooks into the LLVM infrastructure at all. It works with any and all LLVM
passes and code generators, and does not need to "know" how they work. Because
of this, it may appear to do stupid things or miss obvious
simplifications. <tt>bugpoint</tt> is also designed to trade off programmer
time for computer time in the compiler-debugging process; consequently, it may
take a long period of (unattended) time to reduce a test case, but we feel it
is still worth it. Note that <tt>bugpoint</tt> is generally very quick unless
debugging a miscompilation where each test of the program (which requires
executing it) takes a long time.</p>
<!-- ======================================================================= -->
<h3>
<a name="autoselect">Automatic Debugger Selection</a>
</h3>
<div>
<p><tt>bugpoint</tt> reads each <tt>.bc</tt> or <tt>.ll</tt> file specified on
the command line and links them together into a single module, called the test
program. If any LLVM passes are specified on the command line, it runs these
passes on the test program. If any of the passes crash, or if they produce
malformed output (which causes the verifier to abort), <tt>bugpoint</tt> starts
the <a href="#crashdebug">crash debugger</a>.</p>
<p>Otherwise, if the <tt>-output</tt> option was not specified,
<tt>bugpoint</tt> runs the test program with the C backend (which is assumed to
generate good code) to generate a reference output. Once <tt>bugpoint</tt> has
a reference output for the test program, it tries executing it with the
selected code generator. If the selected code generator crashes,
<tt>bugpoint</tt> starts the <a href="#crashdebug">crash debugger</a> on the
code generator. Otherwise, if the resulting output differs from the reference
output, it assumes the difference resulted from a code generator failure, and
starts the <a href="#codegendebug">code generator debugger</a>.</p>
<p>Finally, if the output of the selected code generator matches the reference
output, <tt>bugpoint</tt> runs the test program after all of the LLVM passes
have been applied to it. If its output differs from the reference output, it
assumes the difference resulted from a failure in one of the LLVM passes, and
enters the <a href="#miscompilationdebug">miscompilation debugger</a>.
Otherwise, there is no problem <tt>bugpoint</tt> can debug.</p>
</div>
<!-- ======================================================================= -->
<h3>
<a name="crashdebug">Crash debugger</a>
</h3>
<div>
<p>If an optimizer or code generator crashes, <tt>bugpoint</tt> will try as hard
as it can to reduce the list of passes (for optimizer crashes) and the size of
the test program. First, <tt>bugpoint</tt> figures out which combination of
optimizer passes triggers the bug. This is useful when debugging a problem
exposed by <tt>opt</tt>, for example, because it runs over 38 passes.</p>
<p>Next, <tt>bugpoint</tt> tries removing functions from the test program, to
reduce its size. Usually it is able to reduce a test program to a single
function, when debugging intraprocedural optimizations. Once the number of
functions has been reduced, it attempts to delete various edges in the control
flow graph, to reduce the size of the function as much as possible. Finally,
<tt>bugpoint</tt> deletes any individual LLVM instructions whose absence does
not eliminate the failure. At the end, <tt>bugpoint</tt> should tell you what
passes crash, give you a bitcode file, and give you instructions on how to
reproduce the failure with <tt>opt</tt> or <tt>llc</tt>.</p>
</div>
<!-- ======================================================================= -->
<h3>
<a name="codegendebug">Code generator debugger</a>
</h3>
<div>
<p>The code generator debugger attempts to narrow down the amount of code that
is being miscompiled by the selected code generator. To do this, it takes the
test program and partitions it into two pieces: one piece which it compiles
with the C backend (into a shared object), and one piece which it runs with
either the JIT or the static LLC compiler. It uses several techniques to
reduce the amount of code pushed through the LLVM code generator, to reduce the
potential scope of the problem. After it is finished, it emits two bitcode
files (called "test" [to be compiled with the code generator] and "safe" [to be
compiled with the C backend], respectively), and instructions for reproducing
the problem. The code generator debugger assumes that the C backend produces
good code.</p>
</div>
<!-- ======================================================================= -->
<h3>
<a name="miscompilationdebug">Miscompilation debugger</a>
</h3>
<div>
<p>The miscompilation debugger works similarly to the code generator debugger.
It works by splitting the test program into two pieces, running the
optimizations specified on one piece, linking the two pieces back together, and
then executing the result. It attempts to narrow down the list of passes to
the one (or few) which are causing the miscompilation, then reduce the portion
of the test program which is being miscompiled. The miscompilation debugger
assumes that the selected code generator is working properly.</p>
</div>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="advice">Advice for using bugpoint</a>
</h2>
<!-- *********************************************************************** -->
<div>
<tt>bugpoint</tt> can be a remarkably useful tool, but it sometimes works in
non-obvious ways. Here are some hints and tips:<p>
<ol>
<li>In the code generator and miscompilation debuggers, <tt>bugpoint</tt> only
works with programs that have deterministic output. Thus, if the program
outputs <tt>argv[0]</tt>, the date, time, or any other "random" data,
<tt>bugpoint</tt> may misinterpret differences in these data, when output,
as the result of a miscompilation. Programs should be temporarily modified
to disable outputs that are likely to vary from run to run.
<li>In the code generator and miscompilation debuggers, debugging will go
faster if you manually modify the program or its inputs to reduce the
runtime, but still exhibit the problem.
<li><tt>bugpoint</tt> is extremely useful when working on a new optimization:
it helps track down regressions quickly. To avoid having to relink
<tt>bugpoint</tt> every time you change your optimization however, have
<tt>bugpoint</tt> dynamically load your optimization with the
<tt>-load</tt> option.
<li><p><tt>bugpoint</tt> can generate a lot of output and run for a long period
of time. It is often useful to capture the output of the program to file.
For example, in the C shell, you can run:</p>
<div class="doc_code">
<p><tt>bugpoint ... |&amp; tee bugpoint.log</tt></p>
</div>
<p>to get a copy of <tt>bugpoint</tt>'s output in the file
<tt>bugpoint.log</tt>, as well as on your terminal.</p>
<li><tt>bugpoint</tt> cannot debug problems with the LLVM linker. If
<tt>bugpoint</tt> crashes before you see its "All input ok" message,
you might try <tt>llvm-link -v</tt> on the same set of input files. If
that also crashes, you may be experiencing a linker bug.
<li><tt>bugpoint</tt> is useful for proactively finding bugs in LLVM.
Invoking <tt>bugpoint</tt> with the <tt>-find-bugs</tt> option will cause
the list of specified optimizations to be randomized and applied to the
program. This process will repeat until a bug is found or the user
kills <tt>bugpoint</tt>.
</ol>
</div>
<!-- *********************************************************************** -->
<h2>
<a name="notEnough">What to do when bugpoint isn't enough</a>
</h2>
<!-- *********************************************************************** -->
<div>
<p>Sometimes, <tt>bugpoint</tt> is not enough. In particular, InstCombine and
TargetLowering both have visitor structured code with lots of potential
transformations. If the process of using bugpoint has left you with
still too much code to figure out and the problem seems
to be in instcombine, the following steps may help. These same techniques
are useful with TargetLowering as well.</p>
<p>Turn on <tt>-debug-only=instcombine</tt> and see which transformations
within instcombine are firing by selecting out lines with "<tt>IC</tt>"
in them.</p>
<p>At this point, you have a decision to make. Is the number
of transformations small enough to step through them using a debugger?
If so, then try that.</p>
<p>If there are too many transformations, then a source modification
approach may be helpful.
In this approach, you can modify the source code of instcombine
to disable just those transformations that are being performed on your
test input and perform a binary search over the set of transformations.
One set of places to modify are the "<tt>visit*</tt>" methods of
<tt>InstCombiner</tt> (<I>e.g.</I> <tt>visitICmpInst</tt>) by adding a
"<tt>return false</tt>" as the first line of the method.</p>
<p>If that still doesn't remove enough, then change the caller of
<tt>InstCombiner::DoOneIteration</tt>, <tt>InstCombiner::runOnFunction</tt>
to limit the number of iterations.</p>
<p>You may also find it useful to use "<tt>-stats</tt>" now to see what parts
of instcombine are firing. This can guide where to put additional reporting
code.</p>
<p>At this point, if the amount of transformations is still too large, then
inserting code to limit whether or not to execute the body of the code
in the visit function can be helpful. Add a static counter which is
incremented on every invocation of the function. Then add code which
simply returns false on desired ranges. For example:</p>
<div class="doc_code">
<p><tt>static int calledCount = 0;</tt></p>
<p><tt>calledCount++;</tt></p>
<p><tt>DEBUG(if (calledCount &lt; 212) return false);</tt></p>
<p><tt>DEBUG(if (calledCount &gt; 217) return false);</tt></p>
<p><tt>DEBUG(if (calledCount == 213) return false);</tt></p>
<p><tt>DEBUG(if (calledCount == 214) return false);</tt></p>
<p><tt>DEBUG(if (calledCount == 215) return false);</tt></p>
<p><tt>DEBUG(if (calledCount == 216) return false);</tt></p>
<p><tt>DEBUG(dbgs() &lt;&lt; "visitXOR calledCount: " &lt;&lt; calledCount
&lt;&lt; "\n");</tt></p>
<p><tt>DEBUG(dbgs() &lt;&lt; "I: "; I->dump());</tt></p>
</div>
<p>could be added to <tt>visitXOR</tt> to limit <tt>visitXor</tt> to being
applied only to calls 212 and 217. This is from an actual test case and raises
an important point---a simple binary search may not be sufficient, as
transformations that interact may require isolating more than one call.
In TargetLowering, use <tt>return SDNode();</tt> instead of
<tt>return false;</tt>.</p>
<p>Now that that the number of transformations is down to a manageable
number, try examining the output to see if you can figure out which
transformations are being done. If that can be figured out, then
do the usual debugging. If which code corresponds to the transformation
being performed isn't obvious, set a breakpoint after the call count
based disabling and step through the code. Alternatively, you can use
"printf" style debugging to report waypoints.</p>
</div>
<!-- *********************************************************************** -->
<hr>
<address>
<a href="http://jigsaw.w3.org/css-validator/check/referer"><img
src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
<a href="http://validator.w3.org/check/referer"><img
src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
<a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
<a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
Last modified: $Date$
</address>
</body>
</html>