flywithu/debugging
1
0
Fork 0
mirror of https://github.com/ioacademy-jikim/debugging synced 2025-06-08 08:26:14 +00:00
Code Issues Releases Wiki Activity
debugging/02_day/valgrind/valgrind-3.11.0/docs/html/mc-manual.html
jikim 0b589c7986 first commit
2015-12-13 22:34:58 +09:00

2316 lines
121 KiB
HTML
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
<title>4. Memcheck: a memory error detector</title>
<link rel="stylesheet" type="text/css" href="vg_basic.css">
<meta name="generator" content="DocBook XSL Stylesheets V1.78.1">
<link rel="home" href="index.html" title="Valgrind Documentation">
<link rel="up" href="manual.html" title="Valgrind User Manual">
<link rel="prev" href="manual-core-adv.html" title="3. Using and understanding the Valgrind core: Advanced Topics">
<link rel="next" href="cg-manual.html" title="5. Cachegrind: a cache and branch-prediction profiler">
</head>
<body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF">
<div><table class="nav" width="100%" cellspacing="3" cellpadding="3" border="0" summary="Navigation header"><tr>
<td width="22px" align="center" valign="middle"><a accesskey="p" href="manual-core-adv.html"><img src="images/prev.png" width="18" height="21" border="0" alt="Prev"></a></td>
<td width="25px" align="center" valign="middle"><a accesskey="u" href="manual.html"><img src="images/up.png" width="21" height="18" border="0" alt="Up"></a></td>
<td width="31px" align="center" valign="middle"><a accesskey="h" href="index.html"><img src="images/home.png" width="27" height="20" border="0" alt="Up"></a></td>
<th align="center" valign="middle">Valgrind User Manual</th>
<td width="22px" align="center" valign="middle"><a accesskey="n" href="cg-manual.html"><img src="images/next.png" width="18" height="21" border="0" alt="Next"></a></td>
</tr></table></div>
<div class="chapter">
<div class="titlepage"><div><div><h1 class="title">
<a name="mc-manual"></a>4. Memcheck: a memory error detector</h1></div></div></div>
<div class="toc">
<p><b>Table of Contents</b></p>
<dl class="toc">
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.overview">4.1. Overview</a></span></dt>
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.errormsgs">4.2. Explanation of error messages from Memcheck</a></span></dt>
<dd><dl>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.badrw">4.2.1. Illegal read / Illegal write errors</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.uninitvals">4.2.2. Use of uninitialised values</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.bad-syscall-args">4.2.3. Use of uninitialised or unaddressable values in system
calls</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.badfrees">4.2.4. Illegal frees</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.rudefn">4.2.5. When a heap block is freed with an inappropriate deallocation
function</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.overlap">4.2.6. Overlapping source and destination blocks</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.fishyvalue">4.2.7. Fishy argument values</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.leaks">4.2.8. Memory leak detection</a></span></dt>
</dl></dd>
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.options">4.3. Memcheck Command-Line Options</a></span></dt>
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.suppfiles">4.4. Writing suppression files</a></span></dt>
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.machine">4.5. Details of Memcheck's checking machinery</a></span></dt>
<dd><dl>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.value">4.5.1. Valid-value (V) bits</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.vaddress">4.5.2. Valid-address (A) bits</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.together">4.5.3. Putting it all together</a></span></dt>
</dl></dd>
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.monitor-commands">4.6. Memcheck Monitor Commands</a></span></dt>
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.clientreqs">4.7. Client Requests</a></span></dt>
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.mempools">4.8. Memory Pools: describing and working with custom allocators</a></span></dt>
<dt><span class="sect1"><a href="mc-manual.html#mc-manual.mpiwrap">4.9. Debugging MPI Parallel Programs with Valgrind</a></span></dt>
<dd><dl>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.mpiwrap.build">4.9.1. Building and installing the wrappers</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.mpiwrap.gettingstarted">4.9.2. Getting started</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.mpiwrap.controlling">4.9.3. Controlling the wrapper library</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.mpiwrap.limitations.functions">4.9.4. Functions</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.mpiwrap.limitations.types">4.9.5. Types</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.mpiwrap.writingwrappers">4.9.6. Writing new wrappers</a></span></dt>
<dt><span class="sect2"><a href="mc-manual.html#mc-manual.mpiwrap.whattoexpect">4.9.7. What to expect when using the wrappers</a></span></dt>
</dl></dd>
</dl>
</div>
<p>To use this tool, you may specify <code class="option">--tool=memcheck</code>
on the Valgrind command line. You don't have to, though, since Memcheck
is the default tool.</p>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.overview"></a>4.1. Overview</h2></div></div></div>
<p>Memcheck is a memory error detector. It can detect the following
problems that are common in C and C++ programs.</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>Accessing memory you shouldn't, e.g. overrunning and underrunning
heap blocks, overrunning the top of the stack, and accessing memory after
it has been freed.</p></li>
<li class="listitem"><p>Using undefined values, i.e. values that have not been initialised,
or that have been derived from other undefined values.</p></li>
<li class="listitem"><p>Incorrect freeing of heap memory, such as double-freeing heap
blocks, or mismatched use of
<code class="function">malloc</code>/<code class="computeroutput">new</code>/<code class="computeroutput">new[]</code>
versus
<code class="function">free</code>/<code class="computeroutput">delete</code>/<code class="computeroutput">delete[]</code></p></li>
<li class="listitem"><p>Overlapping <code class="computeroutput">src</code> and
<code class="computeroutput">dst</code> pointers in
<code class="computeroutput">memcpy</code> and related
functions.</p></li>
<li class="listitem"><p>Passing a fishy (presumably negative) value to the
<code class="computeroutput">size</code> parameter of a memory
allocation function.</p></li>
<li class="listitem"><p>Memory leaks.</p></li>
</ul></div>
<p>Problems like these can be difficult to find by other means,
often remaining undetected for long periods, then causing occasional,
difficult-to-diagnose crashes.</p>
</div>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.errormsgs"></a>4.2. Explanation of error messages from Memcheck</h2></div></div></div>
<p>Memcheck issues a range of error messages. This section presents a
quick summary of what error messages mean. The precise behaviour of the
error-checking machinery is described in <a class="xref" href="mc-manual.html#mc-manual.machine" title="4.5. Details of Memcheck's checking machinery">Details of Memcheck's checking machinery</a>.</p>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.badrw"></a>4.2.1. Illegal read / Illegal write errors</h3></div></div></div>
<p>For example:</p>
<pre class="programlisting">
Invalid read of size 4
at 0x40F6BBCC: (within /usr/lib/libpng.so.2.1.0.9)
by 0x40F6B804: (within /usr/lib/libpng.so.2.1.0.9)
by 0x40B07FF4: read_png_image(QImageIO *) (kernel/qpngio.cpp:326)
by 0x40AC751B: QImageIO::read() (kernel/qimage.cpp:3621)
Address 0xBFFFF0E0 is not stack'd, malloc'd or free'd
</pre>
<p>This happens when your program reads or writes memory at a place
which Memcheck reckons it shouldn't. In this example, the program did a
4-byte read at address 0xBFFFF0E0, somewhere within the system-supplied
library libpng.so.2.1.0.9, which was called from somewhere else in the
same library, called from line 326 of <code class="filename">qpngio.cpp</code>,
and so on.</p>
<p>Memcheck tries to establish what the illegal address might relate
to, since that's often useful. So, if it points into a block of memory
which has already been freed, you'll be informed of this, and also where
the block was freed. Likewise, if it should turn out to be just off
the end of a heap block, a common result of off-by-one-errors in
array subscripting, you'll be informed of this fact, and also where the
block was allocated. If you use the <code class="option"><a class="xref" href="manual-core.html#opt.read-var-info">--read-var-info</a></code> option Memcheck will run more slowly
but may give a more detailed description of any illegal address.</p>
<p>In this example, Memcheck can't identify the address. Actually
the address is on the stack, but, for some reason, this is not a valid
stack address -- it is below the stack pointer and that isn't allowed.
In this particular case it's probably caused by GCC generating invalid
code, a known bug in some ancient versions of GCC.</p>
<p>Note that Memcheck only tells you that your program is about to
access memory at an illegal address. It can't stop the access from
happening. So, if your program makes an access which normally would
result in a segmentation fault, you program will still suffer the same
fate -- but you will get a message from Memcheck immediately prior to
this. In this particular example, reading junk on the stack is
non-fatal, and the program stays alive.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.uninitvals"></a>4.2.2. Use of uninitialised values</h3></div></div></div>
<p>For example:</p>
<pre class="programlisting">
Conditional jump or move depends on uninitialised value(s)
at 0x402DFA94: _IO_vfprintf (_itoa.h:49)
by 0x402E8476: _IO_printf (printf.c:36)
by 0x8048472: main (tests/manuel1.c:8)
</pre>
<p>An uninitialised-value use error is reported when your program
uses a value which hasn't been initialised -- in other words, is
undefined. Here, the undefined value is used somewhere inside the
<code class="function">printf</code> machinery of the C library. This error was
reported when running the following small program:</p>
<pre class="programlisting">
int main()
{
int x;
printf ("x = %d\n", x);
}</pre>
<p>It is important to understand that your program can copy around
junk (uninitialised) data as much as it likes. Memcheck observes this
and keeps track of the data, but does not complain. A complaint is
issued only when your program attempts to make use of uninitialised
data in a way that might affect your program's externally-visible behaviour.
In this example, <code class="varname">x</code> is uninitialised. Memcheck observes
the value being passed to <code class="function">_IO_printf</code> and thence to
<code class="function">_IO_vfprintf</code>, but makes no comment. However,
<code class="function">_IO_vfprintf</code> has to examine the value of
<code class="varname">x</code> so it can turn it into the corresponding ASCII string,
and it is at this point that Memcheck complains.</p>
<p>Sources of uninitialised data tend to be:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>Local variables in procedures which have not been initialised,
as in the example above.</p></li>
<li class="listitem"><p>The contents of heap blocks (allocated with
<code class="function">malloc</code>, <code class="function">new</code>, or a similar
function) before you (or a constructor) write something there.
</p></li>
</ul></div>
<p>To see information on the sources of uninitialised data in your
program, use the <code class="option">--track-origins=yes</code> option. This
makes Memcheck run more slowly, but can make it much easier to track down
the root causes of uninitialised value errors.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.bad-syscall-args"></a>4.2.3. Use of uninitialised or unaddressable values in system
calls</h3></div></div></div>
<p>Memcheck checks all parameters to system calls:
</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>It checks all the direct parameters themselves, whether they are
initialised.</p></li>
<li class="listitem"><p>Also, if a system call needs to read from a buffer provided by
your program, Memcheck checks that the entire buffer is addressable
and its contents are initialised.</p></li>
<li class="listitem"><p>Also, if the system call needs to write to a user-supplied
buffer, Memcheck checks that the buffer is addressable.</p></li>
</ul></div>
<p>
</p>
<p>After the system call, Memcheck updates its tracked information to
precisely reflect any changes in memory state caused by the system
call.</p>
<p>Here's an example of two system calls with invalid parameters:</p>
<pre class="programlisting">
#include &lt;stdlib.h&gt;
#include &lt;unistd.h&gt;
int main( void )
{
char* arr = malloc(10);
int* arr2 = malloc(sizeof(int));
write( 1 /* stdout */, arr, 10 );
exit(arr2[0]);
}
</pre>
<p>You get these complaints ...</p>
<pre class="programlisting">
Syscall param write(buf) points to uninitialised byte(s)
at 0x25A48723: __write_nocancel (in /lib/tls/libc-2.3.3.so)
by 0x259AFAD3: __libc_start_main (in /lib/tls/libc-2.3.3.so)
by 0x8048348: (within /auto/homes/njn25/grind/head4/a.out)
Address 0x25AB8028 is 0 bytes inside a block of size 10 alloc'd
at 0x259852B0: malloc (vg_replace_malloc.c:130)
by 0x80483F1: main (a.c:5)
Syscall param exit(error_code) contains uninitialised byte(s)
at 0x25A21B44: __GI__exit (in /lib/tls/libc-2.3.3.so)
by 0x8048426: main (a.c:8)
</pre>
<p>... because the program has (a) written uninitialised junk
from the heap block to the standard output, and (b) passed an
uninitialised value to <code class="function">exit</code>. Note that the first
error refers to the memory pointed to by
<code class="computeroutput">buf</code> (not
<code class="computeroutput">buf</code> itself), but the second error
refers directly to <code class="computeroutput">exit</code>'s argument
<code class="computeroutput">arr2[0]</code>.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.badfrees"></a>4.2.4. Illegal frees</h3></div></div></div>
<p>For example:</p>
<pre class="programlisting">
Invalid free()
at 0x4004FFDF: free (vg_clientmalloc.c:577)
by 0x80484C7: main (tests/doublefree.c:10)
Address 0x3807F7B4 is 0 bytes inside a block of size 177 free'd
at 0x4004FFDF: free (vg_clientmalloc.c:577)
by 0x80484C7: main (tests/doublefree.c:10)
</pre>
<p>Memcheck keeps track of the blocks allocated by your program
with <code class="function">malloc</code>/<code class="computeroutput">new</code>,
so it can know exactly whether or not the argument to
<code class="function">free</code>/<code class="computeroutput">delete</code> is
legitimate or not. Here, this test program has freed the same block
twice. As with the illegal read/write errors, Memcheck attempts to
make sense of the address freed. If, as here, the address is one
which has previously been freed, you wil be told that -- making
duplicate frees of the same block easy to spot. You will also get this
message if you try to free a pointer that doesn't point to the start of a
heap block.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.rudefn"></a>4.2.5. When a heap block is freed with an inappropriate deallocation
function</h3></div></div></div>
<p>In the following example, a block allocated with
<code class="function">new[]</code> has wrongly been deallocated with
<code class="function">free</code>:</p>
<pre class="programlisting">
Mismatched free() / delete / delete []
at 0x40043249: free (vg_clientfuncs.c:171)
by 0x4102BB4E: QGArray::~QGArray(void) (tools/qgarray.cpp:149)
by 0x4C261C41: PptDoc::~PptDoc(void) (include/qmemarray.h:60)
by 0x4C261F0E: PptXml::~PptXml(void) (pptxml.cc:44)
Address 0x4BB292A8 is 0 bytes inside a block of size 64 alloc'd
at 0x4004318C: operator new[](unsigned int) (vg_clientfuncs.c:152)
by 0x4C21BC15: KLaola::readSBStream(int) const (klaola.cc:314)
by 0x4C21C155: KLaola::stream(KLaola::OLENode const *) (klaola.cc:416)
by 0x4C21788F: OLEFilter::convert(QCString const &amp;) (olefilter.cc:272)
</pre>
<p>In <code class="literal">C++</code> it's important to deallocate memory in a
way compatible with how it was allocated. The deal is:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>If allocated with
<code class="function">malloc</code>,
<code class="function">calloc</code>,
<code class="function">realloc</code>,
<code class="function">valloc</code> or
<code class="function">memalign</code>, you must
deallocate with <code class="function">free</code>.</p></li>
<li class="listitem"><p>If allocated with <code class="function">new</code>, you must deallocate
with <code class="function">delete</code>.</p></li>
<li class="listitem"><p>If allocated with <code class="function">new[]</code>, you must
deallocate with <code class="function">delete[]</code>.</p></li>
</ul></div>
<p>The worst thing is that on Linux apparently it doesn't matter if
you do mix these up, but the same program may then crash on a
different platform, Solaris for example. So it's best to fix it
properly. According to the KDE folks "it's amazing how many C++
programmers don't know this".</p>
<p>The reason behind the requirement is as follows. In some C++
implementations, <code class="function">delete[]</code> must be used for
objects allocated by <code class="function">new[]</code> because the compiler
stores the size of the array and the pointer-to-member to the
destructor of the array's content just before the pointer actually
returned. <code class="function">delete</code> doesn't account for this and will get
confused, possibly corrupting the heap.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.overlap"></a>4.2.6. Overlapping source and destination blocks</h3></div></div></div>
<p>The following C library functions copy some data from one
memory block to another (or something similar):
<code class="function">memcpy</code>,
<code class="function">strcpy</code>,
<code class="function">strncpy</code>,
<code class="function">strcat</code>,
<code class="function">strncat</code>.
The blocks pointed to by their <code class="computeroutput">src</code> and
<code class="computeroutput">dst</code> pointers aren't allowed to overlap.
The POSIX standards have wording along the lines "If copying takes place
between objects that overlap, the behavior is undefined." Therefore,
Memcheck checks for this.
</p>
<p>For example:</p>
<pre class="programlisting">
==27492== Source and destination overlap in memcpy(0xbffff294, 0xbffff280, 21)
==27492== at 0x40026CDC: memcpy (mc_replace_strmem.c:71)
==27492== by 0x804865A: main (overlap.c:40)
</pre>
<p>You don't want the two blocks to overlap because one of them could
get partially overwritten by the copying.</p>
<p>You might think that Memcheck is being overly pedantic reporting
this in the case where <code class="computeroutput">dst</code> is less than
<code class="computeroutput">src</code>. For example, the obvious way to
implement <code class="function">memcpy</code> is by copying from the first
byte to the last. However, the optimisation guides of some
architectures recommend copying from the last byte down to the first.
Also, some implementations of <code class="function">memcpy</code> zero
<code class="computeroutput">dst</code> before copying, because zeroing the
destination's cache line(s) can improve performance.</p>
<p>The moral of the story is: if you want to write truly portable
code, don't make any assumptions about the language
implementation.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.fishyvalue"></a>4.2.7. Fishy argument values</h3></div></div></div>
<p>All memory allocation functions take an argument specifying the
size of the memory block that should be allocated. Clearly, the requested
size should be a non-negative value and is typically not excessively large.
For instance, it is extremely unlikly that the size of an allocation
request exceeds 2**63 bytes on a 64-bit machine. It is much more likely that
such a value is the result of an erroneous size calculation and is in effect
a negative value (that just happens to appear excessively large because
the bit pattern is interpreted as an unsigned integer).
Such a value is called a "fishy value".
The <code class="varname">size</code> argument of the following allocation functions
is checked for being fishy:
<code class="function">malloc</code>,
<code class="function">calloc</code>,
<code class="function">realloc</code>,
<code class="function">memalign</code>,
<code class="function">new</code>,
<code class="function">new []</code>.
<code class="function">__builtin_new</code>,
<code class="function">__builtin_vec_new</code>,
For <code class="function">calloc</code> both arguments are being checked.
</p>
<p>For example:</p>
<pre class="programlisting">
==32233== Argument 'size' of function malloc has a fishy (possibly negative) value: -3
==32233== at 0x4C2CFA7: malloc (vg_replace_malloc.c:298)
==32233== by 0x400555: foo (fishy.c:15)
==32233== by 0x400583: main (fishy.c:23)
</pre>
<p>In earlier Valgrind versions those values were being referred to
as "silly arguments" and no back-trace was included.
</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.leaks"></a>4.2.8. Memory leak detection</h3></div></div></div>
<p>Memcheck keeps track of all heap blocks issued in response to
calls to
<code class="function">malloc</code>/<code class="function">new</code> et al.
So when the program exits, it knows which blocks have not been freed.
</p>
<p>If <code class="option">--leak-check</code> is set appropriately, for each
remaining block, Memcheck determines if the block is reachable from pointers
within the root-set. The root-set consists of (a) general purpose registers
of all threads, and (b) initialised, aligned, pointer-sized data words in
accessible client memory, including stacks.</p>
<p>There are two ways a block can be reached. The first is with a
"start-pointer", i.e. a pointer to the start of the block. The second is with
an "interior-pointer", i.e. a pointer to the middle of the block. There are
several ways we know of that an interior-pointer can occur:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>The pointer might have originally been a start-pointer and have been
moved along deliberately (or not deliberately) by the program. In
particular, this can happen if your program uses tagged pointers, i.e.
if it uses the bottom one, two or three bits of a pointer, which are
normally always zero due to alignment, in order to store extra
information.</p></li>
<li class="listitem"><p>It might be a random junk value in memory, entirely unrelated, just
a coincidence.</p></li>
<li class="listitem"><p>It might be a pointer to the inner char array of a C++
<code class="computeroutput">std::string</code>. For example, some
compilers add 3 words at the beginning of the std::string to
store the length, the capacity and a reference count before the
memory containing the array of characters. They return a pointer
just after these 3 words, pointing at the char array.</p></li>
<li class="listitem"><p>Some code might allocate a block of memory, and use the first 8
bytes to store (block size - 8) as a 64bit number.
<code class="computeroutput">sqlite3MemMalloc</code> does this.</p></li>
<li class="listitem"><p>It might be a pointer to an array of C++ objects (which possess
destructors) allocated with <code class="computeroutput">new[]</code>. In
this case, some compilers store a "magic cookie" containing the array
length at the start of the allocated block, and return a pointer to just
past that magic cookie, i.e. an interior-pointer.
See <a class="ulink" href="http://theory.uwinnipeg.ca/gnu/gcc/gxxint_14.html" target="_top">this
page</a> for more information.</p></li>
<li class="listitem"><p>It might be a pointer to an inner part of a C++ object using
multiple inheritance. </p></li>
</ul></div>
<p>You can optionally activate heuristics to use during the leak
search to detect the interior pointers corresponding to
the <code class="computeroutput">stdstring</code>,
<code class="computeroutput">length64</code>,
<code class="computeroutput">newarray</code>
and <code class="computeroutput">multipleinheritance</code> cases. If the
heuristic detects that an interior pointer corresponds to such a case,
the block will be considered as reachable by the interior
pointer. In other words, the interior pointer will be treated
as if it were a start pointer.</p>
<p>With that in mind, consider the nine possible cases described by the
following figure.</p>
<pre class="programlisting">
Pointer chain AAA Leak Case BBB Leak Case
------------- ------------- -------------
(1) RRR ------------&gt; BBB DR
(2) RRR ---&gt; AAA ---&gt; BBB DR IR
(3) RRR BBB DL
(4) RRR AAA ---&gt; BBB DL IL
(5) RRR ------?-----&gt; BBB (y)DR, (n)DL
(6) RRR ---&gt; AAA -?-&gt; BBB DR (y)IR, (n)DL
(7) RRR -?-&gt; AAA ---&gt; BBB (y)DR, (n)DL (y)IR, (n)IL
(8) RRR -?-&gt; AAA -?-&gt; BBB (y)DR, (n)DL (y,y)IR, (n,y)IL, (_,n)DL
(9) RRR AAA -?-&gt; BBB DL (y)IL, (n)DL
Pointer chain legend:
- RRR: a root set node or DR block
- AAA, BBB: heap blocks
- ---&gt;: a start-pointer
- -?-&gt;: an interior-pointer
Leak Case legend:
- DR: Directly reachable
- IR: Indirectly reachable
- DL: Directly lost
- IL: Indirectly lost
- (y)XY: it's XY if the interior-pointer is a real pointer
- (n)XY: it's XY if the interior-pointer is not a real pointer
- (_)XY: it's XY in either case
</pre>
<p>Every possible case can be reduced to one of the above nine. Memcheck
merges some of these cases in its output, resulting in the following four
leak kinds.</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>"Still reachable". This covers cases 1 and 2 (for the BBB blocks)
above. A start-pointer or chain of start-pointers to the block is
found. Since the block is still pointed at, the programmer could, at
least in principle, have freed it before program exit. "Still reachable"
blocks are very common and arguably not a problem. So, by default,
Memcheck won't report such blocks individually.</p></li>
<li class="listitem"><p>"Definitely lost". This covers case 3 (for the BBB blocks) above.
This means that no pointer to the block can be found. The block is
classified as "lost", because the programmer could not possibly have
freed it at program exit, since no pointer to it exists. This is likely
a symptom of having lost the pointer at some earlier point in the
program. Such cases should be fixed by the programmer.</p></li>
<li class="listitem"><p>"Indirectly lost". This covers cases 4 and 9 (for the BBB blocks)
above. This means that the block is lost, not because there are no
pointers to it, but rather because all the blocks that point to it are
themselves lost. For example, if you have a binary tree and the root
node is lost, all its children nodes will be indirectly lost. Because
the problem will disappear if the definitely lost block that caused the
indirect leak is fixed, Memcheck won't report such blocks individually
by default.</p></li>
<li class="listitem"><p>"Possibly lost". This covers cases 5--8 (for the BBB blocks)
above. This means that a chain of one or more pointers to the block has
been found, but at least one of the pointers is an interior-pointer.
This could just be a random value in memory that happens to point into a
block, and so you shouldn't consider this ok unless you know you have
interior-pointers.</p></li>
</ul></div>
<p>(Note: This mapping of the nine possible cases onto four leak kinds is
not necessarily the best way that leaks could be reported; in particular,
interior-pointers are treated inconsistently. It is possible the
categorisation may be improved in the future.)</p>
<p>Furthermore, if suppressions exists for a block, it will be reported
as "suppressed" no matter what which of the above four kinds it belongs
to.</p>
<p>The following is an example leak summary.</p>
<pre class="programlisting">
LEAK SUMMARY:
definitely lost: 48 bytes in 3 blocks.
indirectly lost: 32 bytes in 2 blocks.
possibly lost: 96 bytes in 6 blocks.
still reachable: 64 bytes in 4 blocks.
suppressed: 0 bytes in 0 blocks.
</pre>
<p>If heuristics have been used to consider some blocks as
reachable, the leak summary details the heuristically reachable subset
of 'still reachable:' per heuristic. In the below example, of the 95
bytes still reachable, 87 bytes (56+7+8+16) have been considered
heuristically reachable.
</p>
<pre class="programlisting">
LEAK SUMMARY:
definitely lost: 4 bytes in 1 blocks
indirectly lost: 0 bytes in 0 blocks
possibly lost: 0 bytes in 0 blocks
still reachable: 95 bytes in 6 blocks
of which reachable via heuristic:
stdstring : 56 bytes in 2 blocks
length64 : 16 bytes in 1 blocks
newarray : 7 bytes in 1 blocks
multipleinheritance: 8 bytes in 1 blocks
suppressed: 0 bytes in 0 blocks
</pre>
<p>If <code class="option">--leak-check=full</code> is specified,
Memcheck will give details for each definitely lost or possibly lost block,
including where it was allocated. (Actually, it merges results for all
blocks that have the same leak kind and sufficiently similar stack traces
into a single "loss record". The
<code class="option">--leak-resolution</code> lets you control the
meaning of "sufficiently similar".) It cannot tell you when or how or why
the pointer to a leaked block was lost; you have to work that out for
yourself. In general, you should attempt to ensure your programs do not
have any definitely lost or possibly lost blocks at exit.</p>
<p>For example:</p>
<pre class="programlisting">
8 bytes in 1 blocks are definitely lost in loss record 1 of 14
at 0x........: malloc (vg_replace_malloc.c:...)
by 0x........: mk (leak-tree.c:11)
by 0x........: main (leak-tree.c:39)
88 (8 direct, 80 indirect) bytes in 1 blocks are definitely lost in loss record 13 of 14
at 0x........: malloc (vg_replace_malloc.c:...)
by 0x........: mk (leak-tree.c:11)
by 0x........: main (leak-tree.c:25)
</pre>
<p>The first message describes a simple case of a single 8 byte block
that has been definitely lost. The second case mentions another 8 byte
block that has been definitely lost; the difference is that a further 80
bytes in other blocks are indirectly lost because of this lost block.
The loss records are not presented in any notable order, so the loss record
numbers aren't particularly meaningful. The loss record numbers can be used
in the Valgrind gdbserver to list the addresses of the leaked blocks and/or give
more details about how a block is still reachable.</p>
<p>The option <code class="option">--show-leak-kinds=&lt;set&gt;</code>
controls the set of leak kinds to show
when <code class="option">--leak-check=full</code> is specified. </p>
<p>The <code class="option">&lt;set&gt;</code> of leak kinds is specified
in one of the following ways:
</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>a comma separated list of one or more of
<code class="option">definite indirect possible reachable</code>.</p></li>
<li class="listitem"><p><code class="option">all</code> to specify the complete set (all leak kinds).</p></li>
<li class="listitem"><p><code class="option">none</code> for the empty set.</p></li>
</ul></div>
<p>
</p>
<p> The default value for the leak kinds to show is
<code class="option">--show-leak-kinds=definite,possible</code>.
</p>
<p>To also show the reachable and indirectly lost blocks in
addition to the definitely and possibly lost blocks, you can
use <code class="option">--show-leak-kinds=all</code>. To only show the
reachable and indirectly lost blocks, use
<code class="option">--show-leak-kinds=indirect,reachable</code>. The reachable
and indirectly lost blocks will then be presented as shown in
the following two examples.</p>
<pre class="programlisting">
64 bytes in 4 blocks are still reachable in loss record 2 of 4
at 0x........: malloc (vg_replace_malloc.c:177)
by 0x........: mk (leak-cases.c:52)
by 0x........: main (leak-cases.c:74)
32 bytes in 2 blocks are indirectly lost in loss record 1 of 4
at 0x........: malloc (vg_replace_malloc.c:177)
by 0x........: mk (leak-cases.c:52)
by 0x........: main (leak-cases.c:80)
</pre>
<p>Because there are different kinds of leaks with different
severities, an interesting question is: which leaks should be
counted as true "errors" and which should not?
</p>
<p> The answer to this question affects the numbers printed in
the <code class="computeroutput">ERROR SUMMARY</code> line, and also the
effect of the <code class="option">--error-exitcode</code> option. First, a leak
is only counted as a true "error"
if <code class="option">--leak-check=full</code> is specified. Then, the
option <code class="option">--errors-for-leak-kinds=&lt;set&gt;</code> controls
the set of leak kinds to consider as errors. The default value
is <code class="option">--errors-for-leak-kinds=definite,possible</code>
</p>
</div>
</div>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.options"></a>4.3. Memcheck Command-Line Options</h2></div></div></div>
<div class="variablelist">
<a name="mc.opts.list"></a><dl class="variablelist">
<dt>
<a name="opt.leak-check"></a><span class="term">
<code class="option">--leak-check=&lt;no|summary|yes|full&gt; [default: summary] </code>
</span>
</dt>
<dd><p>When enabled, search for memory leaks when the client
program finishes. If set to <code class="varname">summary</code>, it says how
many leaks occurred. If set to <code class="varname">full</code> or
<code class="varname">yes</code>, each individual leak will be shown
in detail and/or counted as an error, as specified by the options
<code class="option">--show-leak-kinds</code> and
<code class="option">--errors-for-leak-kinds</code>. </p></dd>
<dt>
<a name="opt.leak-resolution"></a><span class="term">
<code class="option">--leak-resolution=&lt;low|med|high&gt; [default: high] </code>
</span>
</dt>
<dd>
<p>When doing leak checking, determines how willing
Memcheck is to consider different backtraces to
be the same for the purposes of merging multiple leaks into a single
leak report. When set to <code class="varname">low</code>, only the first
two entries need match. When <code class="varname">med</code>, four entries
have to match. When <code class="varname">high</code>, all entries need to
match.</p>
<p>For hardcore leak debugging, you probably want to use
<code class="option">--leak-resolution=high</code> together with
<code class="option">--num-callers=40</code> or some such large number.
</p>
<p>Note that the <code class="option">--leak-resolution</code> setting
does not affect Memcheck's ability to find
leaks. It only changes how the results are presented.</p>
</dd>
<dt>
<a name="opt.show-leak-kinds"></a><span class="term">
<code class="option">--show-leak-kinds=&lt;set&gt; [default: definite,possible] </code>
</span>
</dt>
<dd>
<p>Specifies the leak kinds to show in a <code class="varname">full</code>
leak search, in one of the following ways: </p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>a comma separated list of one or more of
<code class="option">definite indirect possible reachable</code>.</p></li>
<li class="listitem"><p><code class="option">all</code> to specify the complete set (all leak kinds).
It is equivalent to
<code class="option">--show-leak-kinds=definite,indirect,possible,reachable</code>.</p></li>
<li class="listitem"><p><code class="option">none</code> for the empty set.</p></li>
</ul></div>
</dd>
<dt>
<a name="opt.errors-for-leak-kinds"></a><span class="term">
<code class="option">--errors-for-leak-kinds=&lt;set&gt; [default: definite,possible] </code>
</span>
</dt>
<dd><p>Specifies the leak kinds to count as errors in a
<code class="varname">full</code> leak search. The
<code class="option">&lt;set&gt;</code> is specified similarly to
<code class="option">--show-leak-kinds</code>
</p></dd>
<dt>
<a name="opt.leak-check-heuristics"></a><span class="term">
<code class="option">--leak-check-heuristics=&lt;set&gt; [default: all] </code>
</span>
</dt>
<dd>
<p>Specifies the set of leak check heuristics to be used
during leak searches. The heuristics control which interior pointers
to a block cause it to be considered as reachable.
The heuristic set is specified in one of the following ways:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>a comma separated list of one or more of
<code class="option">stdstring length64 newarray multipleinheritance</code>.</p></li>
<li class="listitem"><p><code class="option">all</code> to activate the complete set of
heuristics.
It is equivalent to
<code class="option">--leak-check-heuristics=stdstring,length64,newarray,multipleinheritance</code>.</p></li>
<li class="listitem"><p><code class="option">none</code> for the empty set.</p></li>
</ul></div>
</dd>
<dt>
<a name="opt.show-reachable"></a><span class="term">
<code class="option">--show-reachable=&lt;yes|no&gt; </code>
, </span><span class="term">
<code class="option">--show-possibly-lost=&lt;yes|no&gt; </code>
</span>
</dt>
<dd>
<p>These options provide an alternative way to specify the leak kinds to show:
</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>
<code class="option">--show-reachable=no --show-possibly-lost=yes</code> is equivalent to
<code class="option">--show-leak-kinds=definite,possible</code>.
</p></li>
<li class="listitem"><p>
<code class="option">--show-reachable=no --show-possibly-lost=no</code> is equivalent to
<code class="option">--show-leak-kinds=definite</code>.
</p></li>
<li class="listitem"><p>
<code class="option">--show-reachable=yes</code> is equivalent to
<code class="option">--show-leak-kinds=all</code>.
</p></li>
</ul></div>
</dd>
<dt>
<a name="opt.undef-value-errors"></a><span class="term">
<code class="option">--undef-value-errors=&lt;yes|no&gt; [default: yes] </code>
</span>
</dt>
<dd><p>Controls whether Memcheck reports
uses of undefined value errors. Set this to
<code class="varname">no</code> if you don't want to see undefined value
errors. It also has the side effect of speeding up
Memcheck somewhat.
</p></dd>
<dt>
<a name="opt.track-origins"></a><span class="term">
<code class="option">--track-origins=&lt;yes|no&gt; [default: no] </code>
</span>
</dt>
<dd>
<p>Controls whether Memcheck tracks
the origin of uninitialised values. By default, it does not,
which means that although it can tell you that an
uninitialised value is being used in a dangerous way, it
cannot tell you where the uninitialised value came from. This
often makes it difficult to track down the root problem.
</p>
<p>When set
to <code class="varname">yes</code>, Memcheck keeps
track of the origins of all uninitialised values. Then, when
an uninitialised value error is
reported, Memcheck will try to show the
origin of the value. An origin can be one of the following
four places: a heap block, a stack allocation, a client
request, or miscellaneous other sources (eg, a call
to <code class="varname">brk</code>).
</p>
<p>For uninitialised values originating from a heap
block, Memcheck shows where the block was
allocated. For uninitialised values originating from a stack
allocation, Memcheck can tell you which
function allocated the value, but no more than that -- typically
it shows you the source location of the opening brace of the
function. So you should carefully check that all of the
function's local variables are initialised properly.
</p>
<p>Performance overhead: origin tracking is expensive. It
halves Memcheck's speed and increases
memory use by a minimum of 100MB, and possibly more.
Nevertheless it can drastically reduce the effort required to
identify the root cause of uninitialised value errors, and so
is often a programmer productivity win, despite running
more slowly.
</p>
<p>Accuracy: Memcheck tracks origins
quite accurately. To avoid very large space and time
overheads, some approximations are made. It is possible,
although unlikely, that Memcheck will report an incorrect origin, or
not be able to identify any origin.
</p>
<p>Note that the combination
<code class="option">--track-origins=yes</code>
and <code class="option">--undef-value-errors=no</code> is
nonsensical. Memcheck checks for and
rejects this combination at startup.
</p>
</dd>
<dt>
<a name="opt.partial-loads-ok"></a><span class="term">
<code class="option">--partial-loads-ok=&lt;yes|no&gt; [default: yes] </code>
</span>
</dt>
<dd>
<p>Controls how Memcheck handles 32-, 64-, 128- and 256-bit
naturally aligned loads from addresses for which some bytes are
addressable and others are not. When <code class="varname">yes</code>, such
loads do not produce an address error. Instead, loaded bytes
originating from illegal addresses are marked as uninitialised, and
those corresponding to legal addresses are handled in the normal
way.</p>
<p>When <code class="varname">no</code>, loads from partially invalid
addresses are treated the same as loads from completely invalid
addresses: an illegal-address error is issued, and the resulting
bytes are marked as initialised.</p>
<p>Note that code that behaves in this way is in violation of
the ISO C/C++ standards, and should be considered broken. If
at all possible, such code should be fixed.</p>
</dd>
<dt>
<a name="opt.expensive-definedness-checks"></a><span class="term">
<code class="option">--expensive-definedness-checks=&lt;yes|no&gt; [default: no] </code>
</span>
</dt>
<dd><p>Controls whether Memcheck should employ more precise but also more
expensive (time consuming) algorithms when checking the definedness of a
value. The default setting is not to do that and it is usually
sufficient. However, for highly optimised code valgrind may sometimes
incorrectly complain.
Invoking valgrind with <code class="option">--expensive-definedness-checks=yes</code>
helps but comes at a performance cost. Runtime degradation of
25% have been observed but the extra cost depends a lot on the
application at hand.
</p></dd>
<dt>
<a name="opt.keep-stacktraces"></a><span class="term">
<code class="option">--keep-stacktraces=alloc|free|alloc-and-free|alloc-then-free|none [default: alloc-and-free] </code>
</span>
</dt>
<dd>
<p>Controls which stack trace(s) to keep for malloc'd and/or
free'd blocks.
</p>
<p>With <code class="varname">alloc-then-free</code>, a stack trace is
recorded at allocation time, and is associated with the block.
When the block is freed, a second stack trace is recorded, and
this replaces the allocation stack trace. As a result, any "use
after free" errors relating to this block can only show a stack
trace for where the block was freed.
</p>
<p>With <code class="varname">alloc-and-free</code>, both allocation
and the deallocation stack traces for the block are stored.
Hence a "use after free" error will
show both, which may make the error easier to diagnose.
Compared to <code class="varname">alloc-then-free</code>, this setting
slightly increases Valgrind's memory use as the block contains two
references instead of one.
</p>
<p>With <code class="varname">alloc</code>, only the allocation stack
trace is recorded (and reported). With <code class="varname">free</code>,
only the deallocation stack trace is recorded (and reported).
These values somewhat decrease Valgrind's memory and cpu usage.
They can be useful depending on the error types you are
searching for and the level of detail you need to analyse
them. For example, if you are only interested in memory leak
errors, it is sufficient to record the allocation stack traces.
</p>
<p>With <code class="varname">none</code>, no stack traces are recorded
for malloc and free operations. If your program allocates a lot
of blocks and/or allocates/frees from many different stack
traces, this can significantly decrease cpu and/or memory
required. Of course, few details will be reported for errors
related to heap blocks.
</p>
<p>Note that once a stack trace is recorded, Valgrind keeps
the stack trace in memory even if it is not referenced by any
block. Some programs (for example, recursive algorithms) can
generate a huge number of stack traces. If Valgrind uses too
much memory in such circumstances, you can reduce the memory
required with the options <code class="varname">--keep-stacktraces</code>
and/or by using a smaller value for the
option <code class="varname">--num-callers</code>.
</p>
</dd>
<dt>
<a name="opt.freelist-vol"></a><span class="term">
<code class="option">--freelist-vol=&lt;number&gt; [default: 20000000] </code>
</span>
</dt>
<dd>
<p>When the client program releases memory using
<code class="function">free</code> (in <code class="literal">C</code>) or
<code class="computeroutput">delete</code>
(<code class="literal">C++</code>), that memory is not immediately made
available for re-allocation. Instead, it is marked inaccessible
and placed in a queue of freed blocks. The purpose is to defer as
long as possible the point at which freed-up memory comes back
into circulation. This increases the chance that
Memcheck will be able to detect invalid
accesses to blocks for some significant period of time after they
have been freed.</p>
<p>This option specifies the maximum total size, in bytes, of the
blocks in the queue. The default value is twenty million bytes.
Increasing this increases the total amount of memory used by
Memcheck but may detect invalid uses of freed
blocks which would otherwise go undetected.</p>
</dd>
<dt>
<a name="opt.freelist-big-blocks"></a><span class="term">
<code class="option">--freelist-big-blocks=&lt;number&gt; [default: 1000000] </code>
</span>
</dt>
<dd>
<p>When making blocks from the queue of freed blocks available
for re-allocation, Memcheck will in priority re-circulate the blocks
with a size greater or equal to <code class="option">--freelist-big-blocks</code>.
This ensures that freeing big blocks (in particular freeing blocks bigger than
<code class="option">--freelist-vol</code>) does not immediately lead to a re-circulation
of all (or a lot of) the small blocks in the free list. In other words,
this option increases the likelihood to discover dangling pointers
for the "small" blocks, even when big blocks are freed.</p>
<p>Setting a value of 0 means that all the blocks are re-circulated
in a FIFO order. </p>
</dd>
<dt>
<a name="opt.workaround-gcc296-bugs"></a><span class="term">
<code class="option">--workaround-gcc296-bugs=&lt;yes|no&gt; [default: no] </code>
</span>
</dt>
<dd>
<p>When enabled, assume that reads and writes some small
distance below the stack pointer are due to bugs in GCC 2.96, and
does not report them. The "small distance" is 256 bytes by
default. Note that GCC 2.96 is the default compiler on some ancient
Linux distributions (RedHat 7.X) and so you may need to use this
option. Do not use it if you do not have to, as it can cause real
errors to be overlooked. A better alternative is to use a more
recent GCC in which this bug is fixed.</p>
<p>You may also need to use this option when working with
GCC 3.X or 4.X on 32-bit PowerPC Linux. This is because
GCC generates code which occasionally accesses below the
stack pointer, particularly for floating-point to/from integer
conversions. This is in violation of the 32-bit PowerPC ELF
specification, which makes no provision for locations below the
stack pointer to be accessible.</p>
</dd>
<dt>
<a name="opt.show-mismatched-frees"></a><span class="term">
<code class="option">--show-mismatched-frees=&lt;yes|no&gt; [default: yes] </code>
</span>
</dt>
<dd>
<p>When enabled, Memcheck checks that heap blocks are
deallocated using a function that matches the allocating
function. That is, it expects <code class="varname">free</code> to be
used to deallocate blocks allocated
by <code class="varname">malloc</code>, <code class="varname">delete</code> for
blocks allocated by <code class="varname">new</code>,
and <code class="varname">delete[]</code> for blocks allocated
by <code class="varname">new[]</code>. If a mismatch is detected, an
error is reported. This is in general important because in some
environments, freeing with a non-matching function can cause
crashes.</p>
<p>There is however a scenario where such mismatches cannot
be avoided. That is when the user provides implementations of
<code class="varname">new</code>/<code class="varname">new[]</code> that
call <code class="varname">malloc</code> and
of <code class="varname">delete</code>/<code class="varname">delete[]</code> that
call <code class="varname">free</code>, and these functions are
asymmetrically inlined. For example, imagine
that <code class="varname">delete[]</code> is inlined
but <code class="varname">new[]</code> is not. The result is that
Memcheck "sees" all <code class="varname">delete[]</code> calls as direct
calls to <code class="varname">free</code>, even when the program source
contains no mismatched calls.</p>
<p>This causes a lot of confusing and irrelevant error
reports. <code class="varname">--show-mismatched-frees=no</code> disables
these checks. It is not generally advisable to disable them,
though, because you may miss real errors as a result.</p>
</dd>
<dt>
<a name="opt.ignore-ranges"></a><span class="term">
<code class="option">--ignore-ranges=0xPP-0xQQ[,0xRR-0xSS] </code>
</span>
</dt>
<dd><p>Any ranges listed in this option (and multiple ranges can be
specified, separated by commas) will be ignored by Memcheck's
addressability checking.</p></dd>
<dt>
<a name="opt.malloc-fill"></a><span class="term">
<code class="option">--malloc-fill=&lt;hexnumber&gt; </code>
</span>
</dt>
<dd><p>Fills blocks allocated
by <code class="computeroutput">malloc</code>,
<code class="computeroutput">new</code>, etc, but not
by <code class="computeroutput">calloc</code>, with the specified
byte. This can be useful when trying to shake out obscure
memory corruption problems. The allocated area is still
regarded by Memcheck as undefined -- this option only affects its
contents. Note that <code class="option">--malloc-fill</code> does not
affect a block of memory when it is used as argument
to client requests VALGRIND_MEMPOOL_ALLOC or
VALGRIND_MALLOCLIKE_BLOCK.
</p></dd>
<dt>
<a name="opt.free-fill"></a><span class="term">
<code class="option">--free-fill=&lt;hexnumber&gt; </code>
</span>
</dt>
<dd><p>Fills blocks freed
by <code class="computeroutput">free</code>,
<code class="computeroutput">delete</code>, etc, with the
specified byte value. This can be useful when trying to shake out
obscure memory corruption problems. The freed area is still
regarded by Memcheck as not valid for access -- this option only
affects its contents. Note that <code class="option">--free-fill</code> does not
affect a block of memory when it is used as argument to
client requests VALGRIND_MEMPOOL_FREE or VALGRIND_FREELIKE_BLOCK.
</p></dd>
</dl>
</div>
</div>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.suppfiles"></a>4.4. Writing suppression files</h2></div></div></div>
<p>The basic suppression format is described in
<a class="xref" href="manual-core.html#manual-core.suppress" title="2.5. Suppressing errors">Suppressing errors</a>.</p>
<p>The suppression-type (second) line should have the form:</p>
<pre class="programlisting">
Memcheck:suppression_type</pre>
<p>The Memcheck suppression types are as follows:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p><code class="varname">Value1</code>,
<code class="varname">Value2</code>,
<code class="varname">Value4</code>,
<code class="varname">Value8</code>,
<code class="varname">Value16</code>,
meaning an uninitialised-value error when
using a value of 1, 2, 4, 8 or 16 bytes.</p></li>
<li class="listitem"><p><code class="varname">Cond</code> (or its old
name, <code class="varname">Value0</code>), meaning use
of an uninitialised CPU condition code.</p></li>
<li class="listitem"><p><code class="varname">Addr1</code>,
<code class="varname">Addr2</code>,
<code class="varname">Addr4</code>,
<code class="varname">Addr8</code>,
<code class="varname">Addr16</code>,
meaning an invalid address during a
memory access of 1, 2, 4, 8 or 16 bytes respectively.</p></li>
<li class="listitem"><p><code class="varname">Jump</code>, meaning an
jump to an unaddressable location error.</p></li>
<li class="listitem"><p><code class="varname">Param</code>, meaning an
invalid system call parameter error.</p></li>
<li class="listitem"><p><code class="varname">Free</code>, meaning an
invalid or mismatching free.</p></li>
<li class="listitem"><p><code class="varname">Overlap</code>, meaning a
<code class="computeroutput">src</code> /
<code class="computeroutput">dst</code> overlap in
<code class="function">memcpy</code> or a similar function.</p></li>
<li class="listitem"><p><code class="varname">Leak</code>, meaning
a memory leak.</p></li>
</ul></div>
<p><code class="computeroutput">Param</code> errors have a mandatory extra
information line at this point, which is the name of the offending
system call parameter. </p>
<p><code class="computeroutput">Leak</code> errors have an optional
extra information line, with the following format:</p>
<pre class="programlisting">
match-leak-kinds:&lt;set&gt;</pre>
<p>where <code class="computeroutput">&lt;set&gt;</code> specifies which
leak kinds are matched by this suppression entry.
<code class="computeroutput">&lt;set&gt;</code> is specified in the
same way as with the option <code class="option">--show-leak-kinds</code>, that is,
one of the following:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem">a comma separated list of one or more of
<code class="option">definite indirect possible reachable</code>.
</li>
<li class="listitem">
<code class="option">all</code> to specify the complete set (all leak kinds).
</li>
<li class="listitem">
<code class="option">none</code> for the empty set.
</li>
</ul></div>
<p>If this optional extra line is not present, the suppression
entry will match all leak kinds.</p>
<p>Be aware that leak suppressions that are created using
<code class="option">--gen-suppressions</code> will contain this optional extra
line, and therefore may match fewer leaks than you expect. You may
want to remove the line before using the generated
suppressions.</p>
<p>The other Memcheck error kinds do not have extra lines.</p>
<p>
If you give the <code class="option">-v</code> option, Valgrind will print
the list of used suppressions at the end of execution.
For a leak suppression, this output gives the number of different
loss records that match the suppression, and the number of bytes
and blocks suppressed by the suppression.
If the run contains multiple leak checks, the number of bytes and blocks
are reset to zero before each new leak check. Note that the number of different
loss records is not reset to zero.</p>
<p>In the example below, in the last leak search, 7 blocks and 96 bytes have
been suppressed by a suppression with the name
<code class="option">some_leak_suppression</code>:</p>
<pre class="programlisting">
--21041-- used_suppression: 10 some_other_leak_suppression s.supp:14 suppressed: 12,400 bytes in 1 blocks
--21041-- used_suppression: 39 some_leak_suppression s.supp:2 suppressed: 96 bytes in 7 blocks
</pre>
<p>For <code class="varname">ValueN</code> and <code class="varname">AddrN</code>
errors, the first line of the calling context is either the name of
the function in which the error occurred, or, failing that, the full
path of the <code class="filename">.so</code> file or executable containing the
error location. For <code class="varname">Free</code> errors, the first line is
the name of the function doing the freeing (eg,
<code class="function">free</code>, <code class="function">__builtin_vec_delete</code>,
etc). For <code class="varname">Overlap</code> errors, the first line is the name of the
function with the overlapping arguments (eg.
<code class="function">memcpy</code>, <code class="function">strcpy</code>, etc).</p>
<p>The last part of any suppression specifies the rest of the
calling context that needs to be matched.</p>
</div>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.machine"></a>4.5. Details of Memcheck's checking machinery</h2></div></div></div>
<p>Read this section if you want to know, in detail, exactly
what and how Memcheck is checking.</p>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.value"></a>4.5.1. Valid-value (V) bits</h3></div></div></div>
<p>It is simplest to think of Memcheck implementing a synthetic CPU
which is identical to a real CPU, except for one crucial detail. Every
bit (literally) of data processed, stored and handled by the real CPU
has, in the synthetic CPU, an associated "valid-value" bit, which says
whether or not the accompanying bit has a legitimate value. In the
discussions which follow, this bit is referred to as the V (valid-value)
bit.</p>
<p>Each byte in the system therefore has a 8 V bits which follow it
wherever it goes. For example, when the CPU loads a word-size item (4
bytes) from memory, it also loads the corresponding 32 V bits from a
bitmap which stores the V bits for the process' entire address space.
If the CPU should later write the whole or some part of that value to
memory at a different address, the relevant V bits will be stored back
in the V-bit bitmap.</p>
<p>In short, each bit in the system has (conceptually) an associated V
bit, which follows it around everywhere, even inside the CPU. Yes, all the
CPU's registers (integer, floating point, vector and condition registers)
have their own V bit vectors. For this to work, Memcheck uses a great deal
of compression to represent the V bits compactly.</p>
<p>Copying values around does not cause Memcheck to check for, or
report on, errors. However, when a value is used in a way which might
conceivably affect your program's externally-visible behaviour,
the associated V bits are immediately checked. If any of these indicate
that the value is undefined (even partially), an error is reported.</p>
<p>Here's an (admittedly nonsensical) example:</p>
<pre class="programlisting">
int i, j;
int a[10], b[10];
for ( i = 0; i &lt; 10; i++ ) {
j = a[i];
b[i] = j;
}</pre>
<p>Memcheck emits no complaints about this, since it merely copies
uninitialised values from <code class="varname">a[]</code> into
<code class="varname">b[]</code>, and doesn't use them in a way which could
affect the behaviour of the program. However, if
the loop is changed to:</p>
<pre class="programlisting">
for ( i = 0; i &lt; 10; i++ ) {
j += a[i];
}
if ( j == 77 )
printf("hello there\n");
</pre>
<p>then Memcheck will complain, at the
<code class="computeroutput">if</code>, that the condition depends on
uninitialised values. Note that it <span class="command"><strong>doesn't</strong></span> complain
at the <code class="varname">j += a[i];</code>, since at that point the
undefinedness is not "observable". It's only when a decision has to be
made as to whether or not to do the <code class="function">printf</code> -- an
observable action of your program -- that Memcheck complains.</p>
<p>Most low level operations, such as adds, cause Memcheck to use the
V bits for the operands to calculate the V bits for the result. Even if
the result is partially or wholly undefined, it does not
complain.</p>
<p>Checks on definedness only occur in three places: when a value is
used to generate a memory address, when control flow decision needs to
be made, and when a system call is detected, Memcheck checks definedness
of parameters as required.</p>
<p>If a check should detect undefinedness, an error message is
issued. The resulting value is subsequently regarded as well-defined.
To do otherwise would give long chains of error messages. In other
words, once Memcheck reports an undefined value error, it tries to
avoid reporting further errors derived from that same undefined
value.</p>
<p>This sounds overcomplicated. Why not just check all reads from
memory, and complain if an undefined value is loaded into a CPU
register? Well, that doesn't work well, because perfectly legitimate C
programs routinely copy uninitialised values around in memory, and we
don't want endless complaints about that. Here's the canonical example.
Consider a struct like this:</p>
<pre class="programlisting">
struct S { int x; char c; };
struct S s1, s2;
s1.x = 42;
s1.c = 'z';
s2 = s1;
</pre>
<p>The question to ask is: how large is <code class="varname">struct S</code>,
in bytes? An <code class="varname">int</code> is 4 bytes and a
<code class="varname">char</code> one byte, so perhaps a <code class="varname">struct
S</code> occupies 5 bytes? Wrong. All non-toy compilers we know
of will round the size of <code class="varname">struct S</code> up to a whole
number of words, in this case 8 bytes. Not doing this forces compilers
to generate truly appalling code for accessing arrays of
<code class="varname">struct S</code>'s on some architectures.</p>
<p>So <code class="varname">s1</code> occupies 8 bytes, yet only 5 of them will
be initialised. For the assignment <code class="varname">s2 = s1</code>, GCC
generates code to copy all 8 bytes wholesale into <code class="varname">s2</code>
without regard for their meaning. If Memcheck simply checked values as
they came out of memory, it would yelp every time a structure assignment
like this happened. So the more complicated behaviour described above
is necessary. This allows GCC to copy
<code class="varname">s1</code> into <code class="varname">s2</code> any way it likes, and a
warning will only be emitted if the uninitialised values are later
used.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.vaddress"></a>4.5.2. Valid-address (A) bits</h3></div></div></div>
<p>Notice that the previous subsection describes how the validity of
values is established and maintained without having to say whether the
program does or does not have the right to access any particular memory
location. We now consider the latter question.</p>
<p>As described above, every bit in memory or in the CPU has an
associated valid-value (V) bit. In addition, all bytes in memory, but
not in the CPU, have an associated valid-address (A) bit. This
indicates whether or not the program can legitimately read or write that
location. It does not give any indication of the validity of the data
at that location -- that's the job of the V bits -- only whether or not
the location may be accessed.</p>
<p>Every time your program reads or writes memory, Memcheck checks
the A bits associated with the address. If any of them indicate an
invalid address, an error is emitted. Note that the reads and writes
themselves do not change the A bits, only consult them.</p>
<p>So how do the A bits get set/cleared? Like this:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>When the program starts, all the global data areas are
marked as accessible.</p></li>
<li class="listitem"><p>When the program does
<code class="function">malloc</code>/<code class="computeroutput">new</code>,
the A bits for exactly the area allocated, and not a byte more,
are marked as accessible. Upon freeing the area the A bits are
changed to indicate inaccessibility.</p></li>
<li class="listitem"><p>When the stack pointer register (<code class="literal">SP</code>) moves
up or down, A bits are set. The rule is that the area from
<code class="literal">SP</code> up to the base of the stack is marked as
accessible, and below <code class="literal">SP</code> is inaccessible. (If
that sounds illogical, bear in mind that the stack grows down, not
up, on almost all Unix systems, including GNU/Linux.) Tracking
<code class="literal">SP</code> like this has the useful side-effect that the
section of stack used by a function for local variables etc is
automatically marked accessible on function entry and inaccessible
on exit.</p></li>
<li class="listitem"><p>When doing system calls, A bits are changed appropriately.
For example, <code class="literal">mmap</code>
magically makes files appear in the process'
address space, so the A bits must be updated if <code class="literal">mmap</code>
succeeds.</p></li>
<li class="listitem"><p>Optionally, your program can tell Memcheck about such changes
explicitly, using the client request mechanism described
above.</p></li>
</ul></div>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.together"></a>4.5.3. Putting it all together</h3></div></div></div>
<p>Memcheck's checking machinery can be summarised as
follows:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>Each byte in memory has 8 associated V (valid-value) bits,
saying whether or not the byte has a defined value, and a single A
(valid-address) bit, saying whether or not the program currently has
the right to read/write that address. As mentioned above, heavy
use of compression means the overhead is typically around 25%.</p></li>
<li class="listitem"><p>When memory is read or written, the relevant A bits are
consulted. If they indicate an invalid address, Memcheck emits an
Invalid read or Invalid write error.</p></li>
<li class="listitem"><p>When memory is read into the CPU's registers, the relevant V
bits are fetched from memory and stored in the simulated CPU. They
are not consulted.</p></li>
<li class="listitem"><p>When a register is written out to memory, the V bits for that
register are written back to memory too.</p></li>
<li class="listitem"><p>When values in CPU registers are used to generate a memory
address, or to determine the outcome of a conditional branch, the V
bits for those values are checked, and an error emitted if any of
them are undefined.</p></li>
<li class="listitem"><p>When values in CPU registers are used for any other purpose,
Memcheck computes the V bits for the result, but does not check
them.</p></li>
<li class="listitem"><p>Once the V bits for a value in the CPU have been checked, they
are then set to indicate validity. This avoids long chains of
errors.</p></li>
<li class="listitem">
<p>When values are loaded from memory, Memcheck checks the A bits
for that location and issues an illegal-address warning if needed.
In that case, the V bits loaded are forced to indicate Valid,
despite the location being invalid.</p>
<p>This apparently strange choice reduces the amount of confusing
information presented to the user. It avoids the unpleasant
phenomenon in which memory is read from a place which is both
unaddressable and contains invalid values, and, as a result, you get
not only an invalid-address (read/write) error, but also a
potentially large set of uninitialised-value errors, one for every
time the value is used.</p>
<p>There is a hazy boundary case to do with multi-byte loads from
addresses which are partially valid and partially invalid. See
details of the option <code class="option">--partial-loads-ok</code> for details.
</p>
</li>
</ul></div>
<p>Memcheck intercepts calls to <code class="function">malloc</code>,
<code class="function">calloc</code>, <code class="function">realloc</code>,
<code class="function">valloc</code>, <code class="function">memalign</code>,
<code class="function">free</code>, <code class="computeroutput">new</code>,
<code class="computeroutput">new[]</code>,
<code class="computeroutput">delete</code> and
<code class="computeroutput">delete[]</code>. The behaviour you get
is:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p><code class="function">malloc</code>/<code class="function">new</code>/<code class="computeroutput">new[]</code>:
the returned memory is marked as addressable but not having valid
values. This means you have to write to it before you can read
it.</p></li>
<li class="listitem"><p><code class="function">calloc</code>: returned memory is marked both
addressable and valid, since <code class="function">calloc</code> clears
the area to zero.</p></li>
<li class="listitem"><p><code class="function">realloc</code>: if the new size is larger than
the old, the new section is addressable but invalid, as with
<code class="function">malloc</code>. If the new size is smaller, the
dropped-off section is marked as unaddressable. You may only pass to
<code class="function">realloc</code> a pointer previously issued to you by
<code class="function">malloc</code>/<code class="function">calloc</code>/<code class="function">realloc</code>.</p></li>
<li class="listitem"><p><code class="function">free</code>/<code class="computeroutput">delete</code>/<code class="computeroutput">delete[]</code>:
you may only pass to these functions a pointer previously issued
to you by the corresponding allocation function. Otherwise,
Memcheck complains. If the pointer is indeed valid, Memcheck
marks the entire area it points at as unaddressable, and places
the block in the freed-blocks-queue. The aim is to defer as long
as possible reallocation of this block. Until that happens, all
attempts to access it will elicit an invalid-address error, as you
would hope.</p></li>
</ul></div>
</div>
</div>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.monitor-commands"></a>4.6. Memcheck Monitor Commands</h2></div></div></div>
<p>The Memcheck tool provides monitor commands handled by Valgrind's
built-in gdbserver (see <a class="xref" href="manual-core-adv.html#manual-core-adv.gdbserver-commandhandling" title="3.2.5. Monitor command handling by the Valgrind gdbserver">Monitor command handling by the Valgrind gdbserver</a>).
</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem">
<p><code class="varname">xb &lt;addr&gt; [&lt;len&gt;]</code>
shows the definedness (V) bits and values for &lt;len&gt; (default 1)
bytes starting at &lt;addr&gt;.
For each 8 bytes, two lines are output.
</p>
<p>
The first line shows the validity bits for 8 bytes.
The definedness of each byte in the range is given using two hexadecimal
digits. These hexadecimal digits encode the validity of each bit of the
corresponding byte,
using 0 if the bit is defined and 1 if the bit is undefined.
If a byte is not addressable, its validity bits are replaced
by <code class="varname">__</code> (a double underscore).
</p>
<p>
The second line shows the values of the bytes below the corresponding
validity bits. The format used to show the bytes data is similar to the
GDB command 'x /&lt;len&gt;xb &lt;addr&gt;'. The value for a non
addressable bytes is shown as ?? (two question marks).
</p>
<p>
In the following example, <code class="varname">string10</code> is an array
of 10 characters, in which the even numbered bytes are
undefined. In the below example, the byte corresponding
to <code class="varname">string10[5]</code> is not addressable.
</p>
<pre class="programlisting">
(gdb) p &amp;string10
$4 = (char (*)[10]) 0x804a2f0
(gdb) mo xb 0x804a2f0 10
ff 00 ff 00 ff __ ff 00
0x804A2F0: 0x3f 0x6e 0x3f 0x65 0x3f 0x?? 0x3f 0x65
ff 00
0x804A2F8: 0x3f 0x00
Address 0x804A2F0 len 10 has 1 bytes unaddressable
(gdb)
</pre>
<p> The command xb cannot be used with registers. To get
the validity bits of a register, you must start Valgrind with the
option <code class="option">--vgdb-shadow-registers=yes</code>. The validity
bits of a register can then be obtained by printing the 'shadow 1'
corresponding register. In the below x86 example, the register
eax has all its bits undefined, while the register ebx is fully
defined.
</p>
<pre class="programlisting">
(gdb) p /x $eaxs1
$9 = 0xffffffff
(gdb) p /x $ebxs1
$10 = 0x0
(gdb)
</pre>
</li>
<li class="listitem">
<p><code class="varname">get_vbits &lt;addr&gt; [&lt;len&gt;]</code>
shows the definedness (V) bits for &lt;len&gt; (default 1) bytes
starting at &lt;addr&gt; using the same convention as the
<code class="varname">xb</code> command. <code class="varname">get_vbits</code> only
shows the V bits (grouped by 4 bytes). It does not show the values.
If you want to associate V bits with the corresponding byte values, the
<code class="varname">xb</code> command will be easier to use, in particular
on little endian computers when associating undefined parts of an integer
with their V bits values.
</p>
<p>
The following example shows the result of <code class="varname">get_vibts</code>
on the <code class="varname">string10</code> used in the <code class="varname">xb</code>
command explanation.
</p>
<pre class="programlisting">
(gdb) monitor get_vbits 0x804a2f0 10
ff00ff00 ff__ff00 ff00
Address 0x804A2F0 len 10 has 1 bytes unaddressable
(gdb)
</pre>
</li>
<li class="listitem">
<p><code class="varname">make_memory
[noaccess|undefined|defined|Definedifaddressable] &lt;addr&gt;
[&lt;len&gt;]</code> marks the range of &lt;len&gt; (default 1)
bytes at &lt;addr&gt; as having the given status. Parameter
<code class="varname">noaccess</code> marks the range as non-accessible, so
Memcheck will report an error on any access to it.
<code class="varname">undefined</code> or <code class="varname">defined</code> mark
the area as accessible, but Memcheck regards the bytes in it
respectively as having undefined or defined values.
<code class="varname">Definedifaddressable</code> marks as defined, bytes in
the range which are already addressible, but makes no change to
the status of bytes in the range which are not addressible. Note
that the first letter of <code class="varname">Definedifaddressable</code>
is an uppercase D to avoid confusion with <code class="varname">defined</code>.
</p>
<p>
In the following example, the first byte of the
<code class="varname">string10</code> is marked as defined:
</p>
<pre class="programlisting">
(gdb) monitor make_memory defined 0x8049e28 1
(gdb) monitor get_vbits 0x8049e28 10
0000ff00 ff00ff00 ff00
(gdb)
</pre>
</li>
<li class="listitem">
<p><code class="varname">check_memory [addressable|defined] &lt;addr&gt;
[&lt;len&gt;]</code> checks that the range of &lt;len&gt;
(default 1) bytes at &lt;addr&gt; has the specified accessibility.
It then outputs a description of &lt;addr&gt;. In the following
example, a detailed description is available because the
option <code class="option">--read-var-info=yes</code> was given at Valgrind
startup:
</p>
<pre class="programlisting">
(gdb) monitor check_memory defined 0x8049e28 1
Address 0x8049E28 len 1 defined
==14698== Location 0x8049e28 is 0 bytes inside string10[0],
==14698== declared at prog.c:10, in frame #0 of thread 1
(gdb)
</pre>
</li>
<li class="listitem">
<p><code class="varname">leak_check [full*|summary]
[kinds &lt;set&gt;|reachable|possibleleak*|definiteleak]
[heuristics heur1,heur2,...]
[increased*|changed|any]
[unlimited*|limited &lt;max_loss_records_output&gt;]
</code>
performs a leak check. The <code class="varname">*</code> in the arguments
indicates the default values. </p>
<p> If the <code class="varname">[full*|summary]</code> argument is
<code class="varname">summary</code>, only a summary of the leak search is given;
otherwise a full leak report is produced. A full leak report gives
detailed information for each leak: the stack trace where the leaked blocks
were allocated, the number of blocks leaked and their total size. When a
full report is requested, the next two arguments further specify what
kind of leaks to report. A leak's details are shown if they match
both the second and third argument. A full leak report might
output detailed information for many leaks. The nr of leaks for
which information is output can be controlled using
the <code class="varname">limited</code> argument followed by the maximum nr
of leak records to output. If this maximum is reached, the leak
search outputs the records with the biggest number of bytes.
</p>
<p>The <code class="varname">kinds</code> argument controls what kind of blocks
are shown for a <code class="varname">full</code> leak search. The set of leak kinds
to show can be specified using a <code class="varname">&lt;set&gt;</code> similarly
to the command line option <code class="option">--show-leak-kinds</code>.
Alternatively, the value <code class="varname">definiteleak</code>
is equivalent to <code class="varname">kinds definite</code>, the
value <code class="varname">possibleleak</code> is equivalent to
<code class="varname">kinds definite,possible</code> : it will also show
possibly leaked blocks, .i.e those for which only an interior
pointer was found. The value <code class="varname">reachable</code> will
show all block categories (i.e. is equivalent to <code class="varname">kinds
all</code>).
</p>
<p>The <code class="varname">heuristics</code> argument controls the heuristics
used during the leak search. The set of heuristics to use can be specified
using a <code class="varname">&lt;set&gt;</code> similarly
to the command line option <code class="option">--leak-check-heuristics</code>.
The default value for the <code class="varname">heuristics</code> argument is
<code class="varname">heuristics none</code>.
</p>
<p>The <code class="varname">[increased*|changed|any]</code> argument controls what
kinds of changes are shown for a <code class="varname">full</code> leak search. The
value <code class="varname">increased</code> specifies that only block
allocation stacks with an increased number of leaked bytes or
blocks since the previous leak check should be shown. The
value <code class="varname">changed</code> specifies that allocation stacks
with any change since the previous leak check should be shown.
The value <code class="varname">any</code> specifies that all leak entries
should be shown, regardless of any increase or decrease. When
If <code class="varname">increased</code> or <code class="varname">changed</code> are
specified, the leak report entries will show the delta relative to
the previous leak report.
</p>
<p>The following example shows usage of the
<code class="varname">leak_check</code> monitor command on
the <code class="varname">memcheck/tests/leak-cases.c</code> regression
test. The first command outputs one entry having an increase in
the leaked bytes. The second command is the same as the first
command, but uses the abbreviated forms accepted by GDB and the
Valgrind gdbserver. It only outputs the summary information, as
there was no increase since the previous leak search.</p>
<pre class="programlisting">
(gdb) monitor leak_check full possibleleak increased
==19520== 16 (+16) bytes in 1 (+1) blocks are possibly lost in loss record 9 of 12
==19520== at 0x40070B4: malloc (vg_replace_malloc.c:263)
==19520== by 0x80484D5: mk (leak-cases.c:52)
==19520== by 0x804855F: f (leak-cases.c:81)
==19520== by 0x80488E0: main (leak-cases.c:107)
==19520==
==19520== LEAK SUMMARY:
==19520== definitely lost: 32 (+0) bytes in 2 (+0) blocks
==19520== indirectly lost: 16 (+0) bytes in 1 (+0) blocks
==19520== possibly lost: 32 (+16) bytes in 2 (+1) blocks
==19520== still reachable: 96 (+16) bytes in 6 (+1) blocks
==19520== suppressed: 0 (+0) bytes in 0 (+0) blocks
==19520== Reachable blocks (those to which a pointer was found) are not shown.
==19520== To see them, add 'reachable any' args to leak_check
==19520==
(gdb) mo l
==19520== LEAK SUMMARY:
==19520== definitely lost: 32 (+0) bytes in 2 (+0) blocks
==19520== indirectly lost: 16 (+0) bytes in 1 (+0) blocks
==19520== possibly lost: 32 (+0) bytes in 2 (+0) blocks
==19520== still reachable: 96 (+0) bytes in 6 (+0) blocks
==19520== suppressed: 0 (+0) bytes in 0 (+0) blocks
==19520== Reachable blocks (those to which a pointer was found) are not shown.
==19520== To see them, add 'reachable any' args to leak_check
==19520==
(gdb)
</pre>
<p>Note that when using Valgrind's gdbserver, it is not
necessary to rerun
with <code class="option">--leak-check=full</code>
<code class="option">--show-reachable=yes</code> to see the reachable
blocks. You can obtain the same information without rerunning by
using the GDB command <code class="computeroutput">monitor leak_check full
reachable any</code> (or, using
abbreviation: <code class="computeroutput">mo l f r a</code>).
</p>
</li>
<li class="listitem">
<p><code class="varname">block_list &lt;loss_record_nr&gt;|&lt;loss_record_nr_from&gt;..&lt;loss_record_nr_to&gt;
[unlimited*|limited &lt;max_blocks&gt;]
[heuristics heur1,heur2,...]
</code>
shows the list of blocks belonging to
<code class="varname">&lt;loss_record_nr&gt;</code> (or to the loss records range
<code class="varname">&lt;loss_record_nr_from&gt;..&lt;loss_record_nr_to&gt;</code>).
The nr of blocks to print can be controlled using the
<code class="varname">limited</code> argument followed by the maximum nr
of blocks to output.
If one or more heuristics are given, only prints the loss records
and blocks found via one of the given <code class="varname">heur1,heur2,...</code>
heuristics.
</p>
<p> A leak search merges the allocated blocks in loss records :
a loss record re-groups all blocks having the same state (for
example, Definitely Lost) and the same allocation backtrace.
Each loss record is identified in the leak search result
by a loss record number.
The <code class="varname">block_list</code> command shows the loss record information
followed by the addresses and sizes of the blocks which have been
merged in the loss record. If a block was found using an heuristic, the block size
is followed by the heuristic.
</p>
<p> If a directly lost block causes some other blocks to be indirectly
lost, the block_list command will also show these indirectly lost blocks.
The indirectly lost blocks will be indented according to the level of indirection
between the directly lost block and the indirectly lost block(s).
Each indirectly lost block is followed by the reference of its loss record.
</p>
<p> The block_list command can be used on the results of a leak search as long
as no block has been freed after this leak search: as soon as the program frees
a block, a new leak search is needed before block_list can be used again.
</p>
<p>
In the below example, the program leaks a tree structure by losing the pointer to
the block A (top of the tree).
So, the block A is directly lost, causing an indirect
loss of blocks B to G. The first block_list command shows the loss record of A
(a definitely lost block with address 0x4028028, size 16). The addresses and sizes
of the indirectly lost blocks due to block A are shown below the block A.
The second command shows the details of one of the indirect loss records output
by the first command.
</p>
<pre class="programlisting">
A
/ \
B C
/ \ / \
D E F G
</pre>
<pre class="programlisting">
(gdb) bt
#0 main () at leak-tree.c:69
(gdb) monitor leak_check full any
==19552== 112 (16 direct, 96 indirect) bytes in 1 blocks are definitely lost in loss record 7 of 7
==19552== at 0x40070B4: malloc (vg_replace_malloc.c:263)
==19552== by 0x80484D5: mk (leak-tree.c:28)
==19552== by 0x80484FC: f (leak-tree.c:41)
==19552== by 0x8048856: main (leak-tree.c:63)
==19552==
==19552== LEAK SUMMARY:
==19552== definitely lost: 16 bytes in 1 blocks
==19552== indirectly lost: 96 bytes in 6 blocks
==19552== possibly lost: 0 bytes in 0 blocks
==19552== still reachable: 0 bytes in 0 blocks
==19552== suppressed: 0 bytes in 0 blocks
==19552==
(gdb) monitor block_list 7
==19552== 112 (16 direct, 96 indirect) bytes in 1 blocks are definitely lost in loss record 7 of 7
==19552== at 0x40070B4: malloc (vg_replace_malloc.c:263)
==19552== by 0x80484D5: mk (leak-tree.c:28)
==19552== by 0x80484FC: f (leak-tree.c:41)
==19552== by 0x8048856: main (leak-tree.c:63)
==19552== 0x4028028[16]
==19552== 0x4028068[16] indirect loss record 1
==19552== 0x40280E8[16] indirect loss record 3
==19552== 0x4028128[16] indirect loss record 4
==19552== 0x40280A8[16] indirect loss record 2
==19552== 0x4028168[16] indirect loss record 5
==19552== 0x40281A8[16] indirect loss record 6
(gdb) mo b 2
==19552== 16 bytes in 1 blocks are indirectly lost in loss record 2 of 7
==19552== at 0x40070B4: malloc (vg_replace_malloc.c:263)
==19552== by 0x80484D5: mk (leak-tree.c:28)
==19552== by 0x8048519: f (leak-tree.c:43)
==19552== by 0x8048856: main (leak-tree.c:63)
==19552== 0x40280A8[16]
==19552== 0x4028168[16] indirect loss record 5
==19552== 0x40281A8[16] indirect loss record 6
(gdb)
</pre>
</li>
<li class="listitem">
<p><code class="varname">who_points_at &lt;addr&gt; [&lt;len&gt;]</code>
shows all the locations where a pointer to addr is found.
If len is equal to 1, the command only shows the locations pointing
exactly at addr (i.e. the "start pointers" to addr).
If len is &gt; 1, "interior pointers" pointing at the len first bytes
will also be shown.
</p>
<p>The locations searched for are the same as the locations
used in the leak search. So, <code class="varname">who_points_at</code> can a.o.
be used to show why the leak search still can reach a block, or can
search for dangling pointers to a freed block.
Each location pointing at addr (or pointing inside addr if interior pointers
are being searched for) will be described.
</p>
<p>In the below example, the pointers to the 'tree block A' (see example
in command <code class="varname">block_list</code>) is shown before the tree was leaked.
The descriptions are detailed as the option <code class="option">--read-var-info=yes</code>
was given at Valgrind startup. The second call shows the pointers (start and interior
pointers) to block G. The block G (0x40281A8) is reachable via block C (0x40280a8)
and register ECX of tid 1 (tid is the Valgrind thread id).
It is "interior reachable" via the register EBX.
</p>
<pre class="programlisting">
(gdb) monitor who_points_at 0x4028028
==20852== Searching for pointers to 0x4028028
==20852== *0x8049e20 points at 0x4028028
==20852== Location 0x8049e20 is 0 bytes inside global var "t"
==20852== declared at leak-tree.c:35
(gdb) monitor who_points_at 0x40281A8 16
==20852== Searching for pointers pointing in 16 bytes from 0x40281a8
==20852== *0x40280ac points at 0x40281a8
==20852== Address 0x40280ac is 4 bytes inside a block of size 16 alloc'd
==20852== at 0x40070B4: malloc (vg_replace_malloc.c:263)
==20852== by 0x80484D5: mk (leak-tree.c:28)
==20852== by 0x8048519: f (leak-tree.c:43)
==20852== by 0x8048856: main (leak-tree.c:63)
==20852== tid 1 register ECX points at 0x40281a8
==20852== tid 1 register EBX interior points at 2 bytes inside 0x40281a8
(gdb)
</pre>
<p> When <code class="varname">who_points_at</code> finds an interior pointer,
it will report the heuristic(s) with which this interior pointer
will be considered as reachable. Note that this is done independently
of the value of the option <code class="option">--leak-check-heuristics</code>.
In the below example, the loss record 6 indicates a possibly lost
block. <code class="varname">who_points_at</code> reports that there is an interior
pointer pointing in this block, and that the block can be considered
reachable using the heuristic
<code class="computeroutput">multipleinheritance</code>.
</p>
<pre class="programlisting">
(gdb) monitor block_list 6
==3748== 8 bytes in 1 blocks are possibly lost in loss record 6 of 7
==3748== at 0x4007D77: operator new(unsigned int) (vg_replace_malloc.c:313)
==3748== by 0x8048954: main (leak_cpp_interior.cpp:43)
==3748== 0x402A0E0[8]
(gdb) monitor who_points_at 0x402A0E0 8
==3748== Searching for pointers pointing in 8 bytes from 0x402a0e0
==3748== *0xbe8ee078 interior points at 4 bytes inside 0x402a0e0
==3748== Address 0xbe8ee078 is on thread 1's stack
==3748== block at 0x402a0e0 considered reachable by ptr 0x402a0e4 using multipleinheritance heuristic
(gdb)
</pre>
</li>
</ul></div>
</div>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.clientreqs"></a>4.7. Client Requests</h2></div></div></div>
<p>The following client requests are defined in
<code class="filename">memcheck.h</code>.
See <code class="filename">memcheck.h</code> for exact details of their
arguments.</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p><code class="varname">VALGRIND_MAKE_MEM_NOACCESS</code>,
<code class="varname">VALGRIND_MAKE_MEM_UNDEFINED</code> and
<code class="varname">VALGRIND_MAKE_MEM_DEFINED</code>.
These mark address ranges as completely inaccessible,
accessible but containing undefined data, and accessible and
containing defined data, respectively. They return -1, when
run on Valgrind and 0 otherwise.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_MAKE_MEM_DEFINED_IF_ADDRESSABLE</code>.
This is just like <code class="varname">VALGRIND_MAKE_MEM_DEFINED</code> but only
affects those bytes that are already addressable.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_CHECK_MEM_IS_ADDRESSABLE</code> and
<code class="varname">VALGRIND_CHECK_MEM_IS_DEFINED</code>: check immediately
whether or not the given address range has the relevant property,
and if not, print an error message. Also, for the convenience of
the client, returns zero if the relevant property holds; otherwise,
the returned value is the address of the first byte for which the
property is not true. Always returns 0 when not run on
Valgrind.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_CHECK_VALUE_IS_DEFINED</code>: a quick and easy
way to find out whether Valgrind thinks a particular value
(lvalue, to be precise) is addressable and defined. Prints an error
message if not. It has no return value.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_DO_LEAK_CHECK</code>: does a full memory leak
check (like <code class="option">--leak-check=full</code>) right now.
This is useful for incrementally checking for leaks between arbitrary
places in the program's execution. It has no return value.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_DO_ADDED_LEAK_CHECK</code>: same as
<code class="varname"> VALGRIND_DO_LEAK_CHECK</code> but only shows the
entries for which there was an increase in leaked bytes or leaked
number of blocks since the previous leak search. It has no return
value.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_DO_CHANGED_LEAK_CHECK</code>: same as
<code class="varname">VALGRIND_DO_LEAK_CHECK</code> but only shows the
entries for which there was an increase or decrease in leaked
bytes or leaked number of blocks since the previous leak search. It
has no return value.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_DO_QUICK_LEAK_CHECK</code>: like
<code class="varname">VALGRIND_DO_LEAK_CHECK</code>, except it produces only a leak
summary (like <code class="option">--leak-check=summary</code>).
It has no return value.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_COUNT_LEAKS</code>: fills in the four
arguments with the number of bytes of memory found by the previous
leak check to be leaked (i.e. the sum of direct leaks and indirect leaks),
dubious, reachable and suppressed. This is useful in test harness code,
after calling <code class="varname">VALGRIND_DO_LEAK_CHECK</code> or
<code class="varname">VALGRIND_DO_QUICK_LEAK_CHECK</code>.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_COUNT_LEAK_BLOCKS</code>: identical to
<code class="varname">VALGRIND_COUNT_LEAKS</code> except that it returns the
number of blocks rather than the number of bytes in each
category.</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_GET_VBITS</code> and
<code class="varname">VALGRIND_SET_VBITS</code>: allow you to get and set the
V (validity) bits for an address range. You should probably only
set V bits that you have got with
<code class="varname">VALGRIND_GET_VBITS</code>. Only for those who really
know what they are doing.</p></li>
<li class="listitem">
<p><code class="varname">VALGRIND_CREATE_BLOCK</code> and
<code class="varname">VALGRIND_DISCARD</code>. <code class="varname">VALGRIND_CREATE_BLOCK</code>
takes an address, a number of bytes and a character string. The
specified address range is then associated with that string. When
Memcheck reports an invalid access to an address in the range, it
will describe it in terms of this block rather than in terms of
any other block it knows about. Note that the use of this macro
does not actually change the state of memory in any way -- it
merely gives a name for the range.
</p>
<p>At some point you may want Memcheck to stop reporting errors
in terms of the block named
by <code class="varname">VALGRIND_CREATE_BLOCK</code>. To make this
possible, <code class="varname">VALGRIND_CREATE_BLOCK</code> returns a
"block handle", which is a C <code class="varname">int</code> value. You
can pass this block handle to <code class="varname">VALGRIND_DISCARD</code>.
After doing so, Valgrind will no longer relate addressing errors
in the specified range to the block. Passing invalid handles to
<code class="varname">VALGRIND_DISCARD</code> is harmless.
</p>
</li>
</ul></div>
</div>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.mempools"></a>4.8. Memory Pools: describing and working with custom allocators</h2></div></div></div>
<p>Some programs use custom memory allocators, often for performance
reasons. Left to itself, Memcheck is unable to understand the
behaviour of custom allocation schemes as well as it understands the
standard allocators, and so may miss errors and leaks in your program. What
this section describes is a way to give Memcheck enough of a description of
your custom allocator that it can make at least some sense of what is
happening.</p>
<p>There are many different sorts of custom allocator, so Memcheck
attempts to reason about them using a loose, abstract model. We
use the following terminology when describing custom allocation
systems:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p>Custom allocation involves a set of independent "memory pools".
</p></li>
<li class="listitem"><p>Memcheck's notion of a a memory pool consists of a single "anchor
address" and a set of non-overlapping "chunks" associated with the
anchor address.</p></li>
<li class="listitem"><p>Typically a pool's anchor address is the address of a
book-keeping "header" structure.</p></li>
<li class="listitem"><p>Typically the pool's chunks are drawn from a contiguous
"superblock" acquired through the system
<code class="function">malloc</code> or
<code class="function">mmap</code>.</p></li>
</ul></div>
<p>Keep in mind that the last two points above say "typically": the
Valgrind mempool client request API is intentionally vague about the
exact structure of a mempool. There is no specific mention made of
headers or superblocks. Nevertheless, the following picture may help
elucidate the intention of the terms in the API:</p>
<pre class="programlisting">
"pool"
(anchor address)
|
v
+--------+---+
| header | o |
+--------+-|-+
|
v superblock
+------+---+--------------+---+------------------+
| |rzB| allocation |rzB| |
+------+---+--------------+---+------------------+
^ ^
| |
"addr" "addr"+"size"
</pre>
<p>
Note that the header and the superblock may be contiguous or
discontiguous, and there may be multiple superblocks associated with a
single header; such variations are opaque to Memcheck. The API
only requires that your allocation scheme can present sensible values
of "pool", "addr" and "size".</p>
<p>
Typically, before making client requests related to mempools, a client
program will have allocated such a header and superblock for their
mempool, and marked the superblock NOACCESS using the
<code class="varname">VALGRIND_MAKE_MEM_NOACCESS</code> client request.</p>
<p>
When dealing with mempools, the goal is to maintain a particular
invariant condition: that Memcheck believes the unallocated portions
of the pool's superblock (including redzones) are NOACCESS. To
maintain this invariant, the client program must ensure that the
superblock starts out in that state; Memcheck cannot make it so, since
Memcheck never explicitly learns about the superblock of a pool, only
the allocated chunks within the pool.</p>
<p>
Once the header and superblock for a pool are established and properly
marked, there are a number of client requests programs can use to
inform Memcheck about changes to the state of a mempool:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem">
<p>
<code class="varname">VALGRIND_CREATE_MEMPOOL(pool, rzB, is_zeroed)</code>:
This request registers the address <code class="varname">pool</code> as the anchor
address for a memory pool. It also provides a size
<code class="varname">rzB</code>, specifying how large the redzones placed around
chunks allocated from the pool should be. Finally, it provides an
<code class="varname">is_zeroed</code> argument that specifies whether the pool's
chunks are zeroed (more precisely: defined) when allocated.
</p>
<p>
Upon completion of this request, no chunks are associated with the
pool. The request simply tells Memcheck that the pool exists, so that
subsequent calls can refer to it as a pool.
</p>
</li>
<li class="listitem"><p><code class="varname">VALGRIND_DESTROY_MEMPOOL(pool)</code>:
This request tells Memcheck that a pool is being torn down. Memcheck
then removes all records of chunks associated with the pool, as well
as its record of the pool's existence. While destroying its records of
a mempool, Memcheck resets the redzones of any live chunks in the pool
to NOACCESS.
</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_MEMPOOL_ALLOC(pool, addr, size)</code>:
This request informs Memcheck that a <code class="varname">size</code>-byte chunk
has been allocated at <code class="varname">addr</code>, and associates the chunk with the
specified
<code class="varname">pool</code>. If the pool was created with nonzero
<code class="varname">rzB</code> redzones, Memcheck will mark the
<code class="varname">rzB</code> bytes before and after the chunk as NOACCESS. If
the pool was created with the <code class="varname">is_zeroed</code> argument set,
Memcheck will mark the chunk as DEFINED, otherwise Memcheck will mark
the chunk as UNDEFINED.
</p></li>
<li class="listitem"><p><code class="varname">VALGRIND_MEMPOOL_FREE(pool, addr)</code>:
This request informs Memcheck that the chunk at <code class="varname">addr</code>
should no longer be considered allocated. Memcheck will mark the chunk
associated with <code class="varname">addr</code> as NOACCESS, and delete its
record of the chunk's existence.
</p></li>
<li class="listitem">
<p><code class="varname">VALGRIND_MEMPOOL_TRIM(pool, addr, size)</code>:
This request trims the chunks associated with <code class="varname">pool</code>.
The request only operates on chunks associated with
<code class="varname">pool</code>. Trimming is formally defined as:</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: circle; ">
<li class="listitem"><p> All chunks entirely inside the range
<code class="varname">addr..(addr+size-1)</code> are preserved.</p></li>
<li class="listitem"><p>All chunks entirely outside the range
<code class="varname">addr..(addr+size-1)</code> are discarded, as though
<code class="varname">VALGRIND_MEMPOOL_FREE</code> was called on them. </p></li>
<li class="listitem"><p>All other chunks must intersect with the range
<code class="varname">addr..(addr+size-1)</code>; areas outside the
intersection are marked as NOACCESS, as though they had been
independently freed with
<code class="varname">VALGRIND_MEMPOOL_FREE</code>.</p></li>
</ul></div>
<p>This is a somewhat rare request, but can be useful in
implementing the type of mass-free operations common in custom
LIFO allocators.</p>
</li>
<li class="listitem">
<p><code class="varname">VALGRIND_MOVE_MEMPOOL(poolA, poolB)</code>: This
request informs Memcheck that the pool previously anchored at
address <code class="varname">poolA</code> has moved to anchor address
<code class="varname">poolB</code>. This is a rare request, typically only needed
if you <code class="function">realloc</code> the header of a mempool.</p>
<p>No memory-status bits are altered by this request.</p>
</li>
<li class="listitem">
<p>
<code class="varname">VALGRIND_MEMPOOL_CHANGE(pool, addrA, addrB,
size)</code>: This request informs Memcheck that the chunk
previously allocated at address <code class="varname">addrA</code> within
<code class="varname">pool</code> has been moved and/or resized, and should be
changed to cover the region <code class="varname">addrB..(addrB+size-1)</code>. This
is a rare request, typically only needed if you
<code class="function">realloc</code> a superblock or wish to extend a chunk
without changing its memory-status bits.
</p>
<p>No memory-status bits are altered by this request.
</p>
</li>
<li class="listitem"><p><code class="varname">VALGRIND_MEMPOOL_EXISTS(pool)</code>:
This request informs the caller whether or not Memcheck is currently
tracking a mempool at anchor address <code class="varname">pool</code>. It
evaluates to 1 when there is a mempool associated with that address, 0
otherwise. This is a rare request, only useful in circumstances when
client code might have lost track of the set of active mempools.
</p></li>
</ul></div>
</div>
<div class="sect1">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="mc-manual.mpiwrap"></a>4.9. Debugging MPI Parallel Programs with Valgrind</h2></div></div></div>
<p>Memcheck supports debugging of distributed-memory applications
which use the MPI message passing standard. This support consists of a
library of wrapper functions for the
<code class="computeroutput">PMPI_*</code> interface. When incorporated
into the application's address space, either by direct linking or by
<code class="computeroutput">LD_PRELOAD</code>, the wrappers intercept
calls to <code class="computeroutput">PMPI_Send</code>,
<code class="computeroutput">PMPI_Recv</code>, etc. They then
use client requests to inform Memcheck of memory state changes caused
by the function being wrapped. This reduces the number of false
positives that Memcheck otherwise typically reports for MPI
applications.</p>
<p>The wrappers also take the opportunity to carefully check
size and definedness of buffers passed as arguments to MPI functions, hence
detecting errors such as passing undefined data to
<code class="computeroutput">PMPI_Send</code>, or receiving data into a
buffer which is too small.</p>
<p>Unlike most of the rest of Valgrind, the wrapper library is subject to a
BSD-style license, so you can link it into any code base you like.
See the top of <code class="computeroutput">mpi/libmpiwrap.c</code>
for license details.</p>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.mpiwrap.build"></a>4.9.1. Building and installing the wrappers</h3></div></div></div>
<p> The wrapper library will be built automatically if possible.
Valgrind's configure script will look for a suitable
<code class="computeroutput">mpicc</code> to build it with. This must be
the same <code class="computeroutput">mpicc</code> you use to build the
MPI application you want to debug. By default, Valgrind tries
<code class="computeroutput">mpicc</code>, but you can specify a
different one by using the configure-time option
<code class="option">--with-mpicc</code>. Currently the
wrappers are only buildable with
<code class="computeroutput">mpicc</code>s which are based on GNU
GCC or Intel's C++ Compiler.</p>
<p>Check that the configure script prints a line like this:</p>
<pre class="programlisting">
checking for usable MPI2-compliant mpicc and mpi.h... yes, mpicc
</pre>
<p>If it says <code class="computeroutput">... no</code>, your
<code class="computeroutput">mpicc</code> has failed to compile and link
a test MPI2 program.</p>
<p>If the configure test succeeds, continue in the usual way with
<code class="computeroutput">make</code> and <code class="computeroutput">make
install</code>. The final install tree should then contain
<code class="computeroutput">libmpiwrap-&lt;platform&gt;.so</code>.
</p>
<p>Compile up a test MPI program (eg, MPI hello-world) and try
this:</p>
<pre class="programlisting">
LD_PRELOAD=$prefix/lib/valgrind/libmpiwrap-&lt;platform&gt;.so \
mpirun [args] $prefix/bin/valgrind ./hello
</pre>
<p>You should see something similar to the following</p>
<pre class="programlisting">
valgrind MPI wrappers 31901: Active for pid 31901
valgrind MPI wrappers 31901: Try MPIWRAP_DEBUG=help for possible options
</pre>
<p>repeated for every process in the group. If you do not see
these, there is an build/installation problem of some kind.</p>
<p> The MPI functions to be wrapped are assumed to be in an ELF
shared object with soname matching
<code class="computeroutput">libmpi.so*</code>. This is known to be
correct at least for Open MPI and Quadrics MPI, and can easily be
changed if required.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.mpiwrap.gettingstarted"></a>4.9.2. Getting started</h3></div></div></div>
<p>Compile your MPI application as usual, taking care to link it
using the same <code class="computeroutput">mpicc</code> that your
Valgrind build was configured with.</p>
<p>
Use the following basic scheme to run your application on Valgrind with
the wrappers engaged:</p>
<pre class="programlisting">
MPIWRAP_DEBUG=[wrapper-args] \
LD_PRELOAD=$prefix/lib/valgrind/libmpiwrap-&lt;platform&gt;.so \
mpirun [mpirun-args] \
$prefix/bin/valgrind [valgrind-args] \
[application] [app-args]
</pre>
<p>As an alternative to
<code class="computeroutput">LD_PRELOAD</code>ing
<code class="computeroutput">libmpiwrap-&lt;platform&gt;.so</code>, you can
simply link it to your application if desired. This should not disturb
native behaviour of your application in any way.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.mpiwrap.controlling"></a>4.9.3. Controlling the wrapper library</h3></div></div></div>
<p>Environment variable
<code class="computeroutput">MPIWRAP_DEBUG</code> is consulted at
startup. The default behaviour is to print a starting banner</p>
<pre class="programlisting">
valgrind MPI wrappers 16386: Active for pid 16386
valgrind MPI wrappers 16386: Try MPIWRAP_DEBUG=help for possible options
</pre>
<p> and then be relatively quiet.</p>
<p>You can give a list of comma-separated options in
<code class="computeroutput">MPIWRAP_DEBUG</code>. These are</p>
<div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; ">
<li class="listitem"><p><code class="computeroutput">verbose</code>:
show entries/exits of all wrappers. Also show extra
debugging info, such as the status of outstanding
<code class="computeroutput">MPI_Request</code>s resulting
from uncompleted <code class="computeroutput">MPI_Irecv</code>s.</p></li>
<li class="listitem"><p><code class="computeroutput">quiet</code>:
opposite of <code class="computeroutput">verbose</code>, only print
anything when the wrappers want
to report a detected programming error, or in case of catastrophic
failure of the wrappers.</p></li>
<li class="listitem"><p><code class="computeroutput">warn</code>:
by default, functions which lack proper wrappers
are not commented on, just silently
ignored. This causes a warning to be printed for each unwrapped
function used, up to a maximum of three warnings per function.</p></li>
<li class="listitem"><p><code class="computeroutput">strict</code>:
print an error message and abort the program if
a function lacking a wrapper is used.</p></li>
</ul></div>
<p> If you want to use Valgrind's XML output facility
(<code class="option">--xml=yes</code>), you should pass
<code class="computeroutput">quiet</code> in
<code class="computeroutput">MPIWRAP_DEBUG</code> so as to get rid of any
extraneous printing from the wrappers.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.mpiwrap.limitations.functions"></a>4.9.4. Functions</h3></div></div></div>
<p>All MPI2 functions except
<code class="computeroutput">MPI_Wtick</code>,
<code class="computeroutput">MPI_Wtime</code> and
<code class="computeroutput">MPI_Pcontrol</code> have wrappers. The
first two are not wrapped because they return a
<code class="computeroutput">double</code>, which Valgrind's
function-wrap mechanism cannot handle (but it could easily be
extended to do so). <code class="computeroutput">MPI_Pcontrol</code> cannot be
wrapped as it has variable arity:
<code class="computeroutput">int MPI_Pcontrol(const int level, ...)</code></p>
<p>Most functions are wrapped with a default wrapper which does
nothing except complain or abort if it is called, depending on
settings in <code class="computeroutput">MPIWRAP_DEBUG</code> listed
above. The following functions have "real", do-something-useful
wrappers:</p>
<pre class="programlisting">
PMPI_Send PMPI_Bsend PMPI_Ssend PMPI_Rsend
PMPI_Recv PMPI_Get_count
PMPI_Isend PMPI_Ibsend PMPI_Issend PMPI_Irsend
PMPI_Irecv
PMPI_Wait PMPI_Waitall
PMPI_Test PMPI_Testall
PMPI_Iprobe PMPI_Probe
PMPI_Cancel
PMPI_Sendrecv
PMPI_Type_commit PMPI_Type_free
PMPI_Pack PMPI_Unpack
PMPI_Bcast PMPI_Gather PMPI_Scatter PMPI_Alltoall
PMPI_Reduce PMPI_Allreduce PMPI_Op_create
PMPI_Comm_create PMPI_Comm_dup PMPI_Comm_free PMPI_Comm_rank PMPI_Comm_size
PMPI_Error_string
PMPI_Init PMPI_Initialized PMPI_Finalize
</pre>
<p> A few functions such as
<code class="computeroutput">PMPI_Address</code> are listed as
<code class="computeroutput">HAS_NO_WRAPPER</code>. They have no wrapper
at all as there is nothing worth checking, and giving a no-op wrapper
would reduce performance for no reason.</p>
<p> Note that the wrapper library itself can itself generate large
numbers of calls to the MPI implementation, especially when walking
complex types. The most common functions called are
<code class="computeroutput">PMPI_Extent</code>,
<code class="computeroutput">PMPI_Type_get_envelope</code>,
<code class="computeroutput">PMPI_Type_get_contents</code>, and
<code class="computeroutput">PMPI_Type_free</code>. </p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.mpiwrap.limitations.types"></a>4.9.5. Types</h3></div></div></div>
<p> MPI-1.1 structured types are supported, and walked exactly.
The currently supported combiners are
<code class="computeroutput">MPI_COMBINER_NAMED</code>,
<code class="computeroutput">MPI_COMBINER_CONTIGUOUS</code>,
<code class="computeroutput">MPI_COMBINER_VECTOR</code>,
<code class="computeroutput">MPI_COMBINER_HVECTOR</code>
<code class="computeroutput">MPI_COMBINER_INDEXED</code>,
<code class="computeroutput">MPI_COMBINER_HINDEXED</code> and
<code class="computeroutput">MPI_COMBINER_STRUCT</code>. This should
cover all MPI-1.1 types. The mechanism (function
<code class="computeroutput">walk_type</code>) should extend easily to
cover MPI2 combiners.</p>
<p>MPI defines some named structured types
(<code class="computeroutput">MPI_FLOAT_INT</code>,
<code class="computeroutput">MPI_DOUBLE_INT</code>,
<code class="computeroutput">MPI_LONG_INT</code>,
<code class="computeroutput">MPI_2INT</code>,
<code class="computeroutput">MPI_SHORT_INT</code>,
<code class="computeroutput">MPI_LONG_DOUBLE_INT</code>) which are pairs
of some basic type and a C <code class="computeroutput">int</code>.
Unfortunately the MPI specification makes it impossible to look inside
these types and see where the fields are. Therefore these wrappers
assume the types are laid out as <code class="computeroutput">struct { float val;
int loc; }</code> (for
<code class="computeroutput">MPI_FLOAT_INT</code>), etc, and act
accordingly. This appears to be correct at least for Open MPI 1.0.2
and for Quadrics MPI.</p>
<p>If <code class="computeroutput">strict</code> is an option specified
in <code class="computeroutput">MPIWRAP_DEBUG</code>, the application
will abort if an unhandled type is encountered. Otherwise, the
application will print a warning message and continue.</p>
<p>Some effort is made to mark/check memory ranges corresponding to
arrays of values in a single pass. This is important for performance
since asking Valgrind to mark/check any range, no matter how small,
carries quite a large constant cost. This optimisation is applied to
arrays of primitive types (<code class="computeroutput">double</code>,
<code class="computeroutput">float</code>,
<code class="computeroutput">int</code>,
<code class="computeroutput">long</code>, <code class="computeroutput">long
long</code>, <code class="computeroutput">short</code>,
<code class="computeroutput">char</code>, and <code class="computeroutput">long
double</code> on platforms where <code class="computeroutput">sizeof(long
double) == 8</code>). For arrays of all other types, the
wrappers handle each element individually and so there can be a very
large performance cost.</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.mpiwrap.writingwrappers"></a>4.9.6. Writing new wrappers</h3></div></div></div>
<p>
For the most part the wrappers are straightforward. The only
significant complexity arises with nonblocking receives.</p>
<p>The issue is that <code class="computeroutput">MPI_Irecv</code>
states the recv buffer and returns immediately, giving a handle
(<code class="computeroutput">MPI_Request</code>) for the transaction.
Later the user will have to poll for completion with
<code class="computeroutput">MPI_Wait</code> etc, and when the
transaction completes successfully, the wrappers have to paint the
recv buffer. But the recv buffer details are not presented to
<code class="computeroutput">MPI_Wait</code> -- only the handle is. The
library therefore maintains a shadow table which associates
uncompleted <code class="computeroutput">MPI_Request</code>s with the
corresponding buffer address/count/type. When an operation completes,
the table is searched for the associated address/count/type info, and
memory is marked accordingly.</p>
<p>Access to the table is guarded by a (POSIX pthreads) lock, so as
to make the library thread-safe.</p>
<p>The table is allocated with
<code class="computeroutput">malloc</code> and never
<code class="computeroutput">free</code>d, so it will show up in leak
checks.</p>
<p>Writing new wrappers should be fairly easy. The source file is
<code class="computeroutput">mpi/libmpiwrap.c</code>. If possible,
find an existing wrapper for a function of similar behaviour to the
one you want to wrap, and use it as a starting point. The wrappers
are organised in sections in the same order as the MPI 1.1 spec, to
aid navigation. When adding a wrapper, remember to comment out the
definition of the default wrapper in the long list of defaults at the
bottom of the file (do not remove it, just comment it out).</p>
</div>
<div class="sect2">
<div class="titlepage"><div><div><h3 class="title">
<a name="mc-manual.mpiwrap.whattoexpect"></a>4.9.7. What to expect when using the wrappers</h3></div></div></div>
<p>The wrappers should reduce Memcheck's false-error rate on MPI
applications. Because the wrapping is done at the MPI interface,
there will still potentially be a large number of errors reported in
the MPI implementation below the interface. The best you can do is
try to suppress them.</p>
<p>You may also find that the input-side (buffer
length/definedness) checks find errors in your MPI use, for example
passing too short a buffer to
<code class="computeroutput">MPI_Recv</code>.</p>
<p>Functions which are not wrapped may increase the false
error rate. A possible approach is to run with
<code class="computeroutput">MPI_DEBUG</code> containing
<code class="computeroutput">warn</code>. This will show you functions
which lack proper wrappers but which are nevertheless used. You can
then write wrappers for them.
</p>
<p>A known source of potential false errors are the
<code class="computeroutput">PMPI_Reduce</code> family of functions, when
using a custom (user-defined) reduction function. In a reduction
operation, each node notionally sends data to a "central point" which
uses the specified reduction function to merge the data items into a
single item. Hence, in general, data is passed between nodes and fed
to the reduction function, but the wrapper library cannot mark the
transferred data as initialised before it is handed to the reduction
function, because all that happens "inside" the
<code class="computeroutput">PMPI_Reduce</code> call. As a result you
may see false positives reported in your reduction function.</p>
</div>
</div>
</div>
<div>
<br><table class="nav" width="100%" cellspacing="3" cellpadding="2" border="0" summary="Navigation footer">
<tr>
<td rowspan="2" width="40%" align="left">
<a accesskey="p" href="manual-core-adv.html">&lt;&lt; 3. Using and understanding the Valgrind core: Advanced Topics</a> </td>
<td width="20%" align="center"><a accesskey="u" href="manual.html">Up</a></td>
<td rowspan="2" width="40%" align="right"> <a accesskey="n" href="cg-manual.html">5. Cachegrind: a cache and branch-prediction profiler &gt;&gt;</a>
</td>
</tr>
<tr><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td></tr>
</table>
</div>
</body>
</html>
Reference in New Issue View Git Blame Copy Permalink