xref: /netbsd-src/external/bsd/jemalloc.old/dist/doc/jemalloc.html (revision 8e33eff89e26cf71871ead62f0d5063e1313c33a)
1<?xml version="1.0" encoding="utf-8"?>
2<html><head><title>JEMALLOC</title><meta name="generator" content="DocBook XSL Stylesheets V1.78.1"/></head><body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"><div class="refentry"><a name="idm140689766630352"/><div class="titlepage"/><div class="refnamediv"><h2>Name</h2><p>jemalloc — general purpose memory allocation functions</p></div><div class="refsect1"><a name="library"/><h2>LIBRARY</h2><p>This manual describes jemalloc 5.1.0-0-g61efbda7098de6fe64c362d309824864308c36d4.  More information
3    can be found at the <a class="ulink" href="http://jemalloc.net/" target="_top">jemalloc website</a>.</p></div><div class="refsynopsisdiv"><h2>SYNOPSIS</h2><div class="funcsynopsis"><pre class="funcsynopsisinfo">#include &lt;<code class="filename">jemalloc/jemalloc.h</code>&gt;</pre><div class="refsect2"><a name="idm140689767408528"/><h3>Standard API</h3><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">malloc</b>(</code></td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">calloc</b>(</code></td><td>size_t <var class="pdparam">number</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">posix_memalign</b>(</code></td><td>void **<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">aligned_alloc</b>(</code></td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">realloc</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">free</b>(</code></td><td>void *<var class="pdparam">ptr</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="refsect2"><a name="idm140689770688880"/><h3>Non-standard API</h3><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">mallocx</b>(</code></td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void *<b class="fsfunc">rallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">xallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">extra</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">sallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">dallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">sdallocx</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">nallocx</b>(</code></td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctl</b>(</code></td><td>const char *<var class="pdparam">name</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">oldp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">oldlenp</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">newp</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">newlen</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctlnametomib</b>(</code></td><td>const char *<var class="pdparam">name</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">mibp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">miblenp</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">int <b class="fsfunc">mallctlbymib</b>(</code></td><td>const size_t *<var class="pdparam">mib</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">miblen</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">oldp</var>, </td></tr><tr><td> </td><td>size_t *<var class="pdparam">oldlenp</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">newp</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">newlen</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">malloc_stats_print</b>(</code></td><td>void <var class="pdparam">(*write_cb)</var>
4            <code>(</code>void *, const char *<code>)</code>
5          , </td></tr><tr><td> </td><td>void *<var class="pdparam">cbopaque</var>, </td></tr><tr><td> </td><td>const char *<var class="pdparam">opts</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">size_t <b class="fsfunc">malloc_usable_size</b>(</code></td><td>const void *<var class="pdparam">ptr</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">void <b class="fsfunc">(*malloc_message)</b>(</code></td><td>void *<var class="pdparam">cbopaque</var>, </td></tr><tr><td> </td><td>const char *<var class="pdparam">s</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div><p><span class="type">const char *</span><code class="varname">malloc_conf</code>;</p></div></div></div><div class="refsect1"><a name="description"/><h2>DESCRIPTION</h2><div class="refsect2"><a name="idm140689765494448"/><h3>Standard API</h3><p>The <code class="function">malloc()</code> function allocates
6      <em class="parameter"><code>size</code></em> bytes of uninitialized memory.  The allocated
7      space is suitably aligned (after possible pointer coercion) for storage
8      of any type of object.</p><p>The <code class="function">calloc()</code> function allocates
9      space for <em class="parameter"><code>number</code></em> objects, each
10      <em class="parameter"><code>size</code></em> bytes in length.  The result is identical to
11      calling <code class="function">malloc()</code> with an argument of
12      <em class="parameter"><code>number</code></em> * <em class="parameter"><code>size</code></em>, with the
13      exception that the allocated memory is explicitly initialized to zero
14      bytes.</p><p>The <code class="function">posix_memalign()</code> function
15      allocates <em class="parameter"><code>size</code></em> bytes of memory such that the
16      allocation's base address is a multiple of
17      <em class="parameter"><code>alignment</code></em>, and returns the allocation in the value
18      pointed to by <em class="parameter"><code>ptr</code></em>.  The requested
19      <em class="parameter"><code>alignment</code></em> must be a power of 2 at least as large as
20      <code class="code">sizeof(<span class="type">void *</span>)</code>.</p><p>The <code class="function">aligned_alloc()</code> function
21      allocates <em class="parameter"><code>size</code></em> bytes of memory such that the
22      allocation's base address is a multiple of
23      <em class="parameter"><code>alignment</code></em>.  The requested
24      <em class="parameter"><code>alignment</code></em> must be a power of 2.  Behavior is
25      undefined if <em class="parameter"><code>size</code></em> is not an integral multiple of
26      <em class="parameter"><code>alignment</code></em>.</p><p>The <code class="function">realloc()</code> function changes the
27      size of the previously allocated memory referenced by
28      <em class="parameter"><code>ptr</code></em> to <em class="parameter"><code>size</code></em> bytes.  The
29      contents of the memory are unchanged up to the lesser of the new and old
30      sizes.  If the new size is larger, the contents of the newly allocated
31      portion of the memory are undefined.  Upon success, the memory referenced
32      by <em class="parameter"><code>ptr</code></em> is freed and a pointer to the newly
33      allocated memory is returned.  Note that
34      <code class="function">realloc()</code> may move the memory allocation,
35      resulting in a different return value than <em class="parameter"><code>ptr</code></em>.
36      If <em class="parameter"><code>ptr</code></em> is <code class="constant">NULL</code>, the
37      <code class="function">realloc()</code> function behaves identically to
38      <code class="function">malloc()</code> for the specified size.</p><p>The <code class="function">free()</code> function causes the
39      allocated memory referenced by <em class="parameter"><code>ptr</code></em> to be made
40      available for future allocations.  If <em class="parameter"><code>ptr</code></em> is
41      <code class="constant">NULL</code>, no action occurs.</p></div><div class="refsect2"><a name="idm140689765469472"/><h3>Non-standard API</h3><p>The <code class="function">mallocx()</code>,
42      <code class="function">rallocx()</code>,
43      <code class="function">xallocx()</code>,
44      <code class="function">sallocx()</code>,
45      <code class="function">dallocx()</code>,
46      <code class="function">sdallocx()</code>, and
47      <code class="function">nallocx()</code> functions all have a
48      <em class="parameter"><code>flags</code></em> argument that can be used to specify
49      options.  The functions only check the options that are contextually
50      relevant.  Use bitwise or (<code class="code">|</code>) operations to
51      specify one or more of the following:
52        </p><div class="variablelist"><dl class="variablelist"><dt><a name="MALLOCX_LG_ALIGN"/><span class="term"><code class="constant">MALLOCX_LG_ALIGN(<em class="parameter"><code>la</code></em>)
53            </code></span></dt><dd><p>Align the memory allocation to start at an address
54            that is a multiple of <code class="code">(1 &lt;&lt;
55            <em class="parameter"><code>la</code></em>)</code>.  This macro does not validate
56            that <em class="parameter"><code>la</code></em> is within the valid
57            range.</p></dd><dt><a name="MALLOCX_ALIGN"/><span class="term"><code class="constant">MALLOCX_ALIGN(<em class="parameter"><code>a</code></em>)
58            </code></span></dt><dd><p>Align the memory allocation to start at an address
59            that is a multiple of <em class="parameter"><code>a</code></em>, where
60            <em class="parameter"><code>a</code></em> is a power of two.  This macro does not
61            validate that <em class="parameter"><code>a</code></em> is a power of 2.
62            </p></dd><dt><a name="MALLOCX_ZERO"/><span class="term"><code class="constant">MALLOCX_ZERO</code></span></dt><dd><p>Initialize newly allocated memory to contain zero
63            bytes.  In the growing reallocation case, the real size prior to
64            reallocation defines the boundary between untouched bytes and those
65            that are initialized to contain zero bytes.  If this macro is
66            absent, newly allocated memory is uninitialized.</p></dd><dt><a name="MALLOCX_TCACHE"/><span class="term"><code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)
67            </code></span></dt><dd><p>Use the thread-specific cache (tcache) specified by
68            the identifier <em class="parameter"><code>tc</code></em>, which must have been
69            acquired via the <a class="link" href="#tcache.create"><quote><code class="mallctl">tcache.create</code></quote></a>
70            mallctl.  This macro does not validate that
71            <em class="parameter"><code>tc</code></em> specifies a valid
72            identifier.</p></dd><dt><a name="MALLOC_TCACHE_NONE"/><span class="term"><code class="constant">MALLOCX_TCACHE_NONE</code></span></dt><dd><p>Do not use a thread-specific cache (tcache).  Unless
73            <code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)</code> or
74            <code class="constant">MALLOCX_TCACHE_NONE</code> is specified, an
75            automatically managed tcache will be used under many circumstances.
76            This macro cannot be used in the same <em class="parameter"><code>flags</code></em>
77            argument as
78            <code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)</code>.</p></dd><dt><a name="MALLOCX_ARENA"/><span class="term"><code class="constant">MALLOCX_ARENA(<em class="parameter"><code>a</code></em>)
79            </code></span></dt><dd><p>Use the arena specified by the index
80            <em class="parameter"><code>a</code></em>.  This macro has no effect for regions that
81            were allocated via an arena other than the one specified.  This
82            macro does not validate that <em class="parameter"><code>a</code></em> specifies an
83            arena index in the valid range.</p></dd></dl></div><p>
84      </p><p>The <code class="function">mallocx()</code> function allocates at
85      least <em class="parameter"><code>size</code></em> bytes of memory, and returns a pointer
86      to the base address of the allocation.  Behavior is undefined if
87      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>.</p><p>The <code class="function">rallocx()</code> function resizes the
88      allocation at <em class="parameter"><code>ptr</code></em> to be at least
89      <em class="parameter"><code>size</code></em> bytes, and returns a pointer to the base
90      address of the resulting allocation, which may or may not have moved from
91      its original location.  Behavior is undefined if
92      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>.</p><p>The <code class="function">xallocx()</code> function resizes the
93      allocation at <em class="parameter"><code>ptr</code></em> in place to be at least
94      <em class="parameter"><code>size</code></em> bytes, and returns the real size of the
95      allocation.  If <em class="parameter"><code>extra</code></em> is non-zero, an attempt is
96      made to resize the allocation to be at least <code class="code">(<em class="parameter"><code>size</code></em> +
97      <em class="parameter"><code>extra</code></em>)</code> bytes, though inability to allocate
98      the extra byte(s) will not by itself result in failure to resize.
99      Behavior is undefined if <em class="parameter"><code>size</code></em> is
100      <code class="constant">0</code>, or if <code class="code">(<em class="parameter"><code>size</code></em> + <em class="parameter"><code>extra</code></em>
101      &gt; <code class="constant">SIZE_T_MAX</code>)</code>.</p><p>The <code class="function">sallocx()</code> function returns the
102      real size of the allocation at <em class="parameter"><code>ptr</code></em>.</p><p>The <code class="function">dallocx()</code> function causes the
103      memory referenced by <em class="parameter"><code>ptr</code></em> to be made available for
104      future allocations.</p><p>The <code class="function">sdallocx()</code> function is an
105      extension of <code class="function">dallocx()</code> with a
106      <em class="parameter"><code>size</code></em> parameter to allow the caller to pass in the
107      allocation size as an optimization.  The minimum valid input size is the
108      original requested size of the allocation, and the maximum valid input
109      size is the corresponding value returned by
110      <code class="function">nallocx()</code> or
111      <code class="function">sallocx()</code>.</p><p>The <code class="function">nallocx()</code> function allocates no
112      memory, but it performs the same size computation as the
113      <code class="function">mallocx()</code> function, and returns the real
114      size of the allocation that would result from the equivalent
115      <code class="function">mallocx()</code> function call, or
116      <code class="constant">0</code> if the inputs exceed the maximum supported size
117      class and/or alignment.  Behavior is undefined if
118      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>.</p><p>The <code class="function">mallctl()</code> function provides a
119      general interface for introspecting the memory allocator, as well as
120      setting modifiable parameters and triggering actions.  The
121      period-separated <em class="parameter"><code>name</code></em> argument specifies a
122      location in a tree-structured namespace; see the <a class="xref" href="#mallctl_namespace" title="MALLCTL NAMESPACE">MALLCTL NAMESPACE</a> section for
123      documentation on the tree contents.  To read a value, pass a pointer via
124      <em class="parameter"><code>oldp</code></em> to adequate space to contain the value, and a
125      pointer to its length via <em class="parameter"><code>oldlenp</code></em>; otherwise pass
126      <code class="constant">NULL</code> and <code class="constant">NULL</code>.  Similarly, to
127      write a value, pass a pointer to the value via
128      <em class="parameter"><code>newp</code></em>, and its length via
129      <em class="parameter"><code>newlen</code></em>; otherwise pass <code class="constant">NULL</code>
130      and <code class="constant">0</code>.</p><p>The <code class="function">mallctlnametomib()</code> function
131      provides a way to avoid repeated name lookups for applications that
132      repeatedly query the same portion of the namespace, by translating a name
133      to a <span class="quote">“<span class="quote">Management Information Base</span>”</span> (MIB) that can be passed
134      repeatedly to <code class="function">mallctlbymib()</code>.  Upon
135      successful return from <code class="function">mallctlnametomib()</code>,
136      <em class="parameter"><code>mibp</code></em> contains an array of
137      <em class="parameter"><code>*miblenp</code></em> integers, where
138      <em class="parameter"><code>*miblenp</code></em> is the lesser of the number of components
139      in <em class="parameter"><code>name</code></em> and the input value of
140      <em class="parameter"><code>*miblenp</code></em>.  Thus it is possible to pass a
141      <em class="parameter"><code>*miblenp</code></em> that is smaller than the number of
142      period-separated name components, which results in a partial MIB that can
143      be used as the basis for constructing a complete MIB.  For name
144      components that are integers (e.g. the 2 in
145      <a class="link" href="#arenas.bin.i.size"><quote><code class="mallctl">arenas.bin.2.size</code></quote></a>),
146      the corresponding MIB component will always be that integer.  Therefore,
147      it is legitimate to construct code like the following: </p><pre class="programlisting">
148unsigned nbins, i;
149size_t mib[4];
150size_t len, miblen;
151
152len = sizeof(nbins);
153mallctl("arenas.nbins", &amp;nbins, &amp;len, NULL, 0);
154
155miblen = 4;
156mallctlnametomib("arenas.bin.0.size", mib, &amp;miblen);
157for (i = 0; i &lt; nbins; i++) {
158	size_t bin_size;
159
160	mib[2] = i;
161	len = sizeof(bin_size);
162	mallctlbymib(mib, miblen, (void *)&amp;bin_size, &amp;len, NULL, 0);
163	/* Do something with bin_size... */
164}</pre><dt><a name="malloc_stats_print_opts"/></dt><dd/><p>The <code class="function">malloc_stats_print()</code> function writes
165      summary statistics via the <em class="parameter"><code>write_cb</code></em> callback
166      function pointer and <em class="parameter"><code>cbopaque</code></em> data passed to
167      <em class="parameter"><code>write_cb</code></em>, or <code class="function">malloc_message()</code>
168      if <em class="parameter"><code>write_cb</code></em> is <code class="constant">NULL</code>.  The
169      statistics are presented in human-readable form unless <span class="quote">“<span class="quote">J</span>”</span> is
170      specified as a character within the <em class="parameter"><code>opts</code></em> string, in
171      which case the statistics are presented in <a class="ulink" href="http://www.json.org/" target="_top">JSON format</a>.  This function can be
172      called repeatedly.  General information that never changes during
173      execution can be omitted by specifying <span class="quote">“<span class="quote">g</span>”</span> as a character
174      within the <em class="parameter"><code>opts</code></em> string.  Note that
175      <code class="function">malloc_message()</code> uses the
176      <code class="function">mallctl*()</code> functions internally, so inconsistent
177      statistics can be reported if multiple threads use these functions
178      simultaneously.  If <code class="option">--enable-stats</code> is specified during
179      configuration, <span class="quote">“<span class="quote">m</span>”</span>, <span class="quote">“<span class="quote">d</span>”</span>, and <span class="quote">“<span class="quote">a</span>”</span>
180      can be specified to omit merged arena, destroyed merged arena, and per
181      arena statistics, respectively; <span class="quote">“<span class="quote">b</span>”</span> and <span class="quote">“<span class="quote">l</span>”</span> can
182      be specified to omit per size class statistics for bins and large objects,
183      respectively; <span class="quote">“<span class="quote">x</span>”</span> can be specified to omit all mutex
184      statistics.  Unrecognized characters are silently ignored.  Note that
185      thread caching may prevent some statistics from being completely up to
186      date, since extra locking would be required to merge counters that track
187      thread cache operations.</p><p>The <code class="function">malloc_usable_size()</code> function
188      returns the usable size of the allocation pointed to by
189      <em class="parameter"><code>ptr</code></em>.  The return value may be larger than the size
190      that was requested during allocation.  The
191      <code class="function">malloc_usable_size()</code> function is not a
192      mechanism for in-place <code class="function">realloc()</code>; rather
193      it is provided solely as a tool for introspection purposes.  Any
194      discrepancy between the requested allocation size and the size reported
195      by <code class="function">malloc_usable_size()</code> should not be
196      depended on, since such behavior is entirely implementation-dependent.
197      </p></div></div><div class="refsect1"><a name="tuning"/><h2>TUNING</h2><p>Once, when the first call is made to one of the memory allocation
198    routines, the allocator initializes its internals based in part on various
199    options that can be specified at compile- or run-time.</p><p>The string specified via <code class="option">--with-malloc-conf</code>, the
200    string pointed to by the global variable <code class="varname">malloc_conf</code>, the
201    <span class="quote">“<span class="quote">name</span>”</span> of the file referenced by the symbolic link named
202    <code class="filename">/etc/malloc.conf</code>, and the value of the
203    environment variable <code class="envar">MALLOC_CONF</code>, will be interpreted, in
204    that order, from left to right as options.  Note that
205    <code class="varname">malloc_conf</code> may be read before
206    <code class="function">main()</code> is entered, so the declaration of
207    <code class="varname">malloc_conf</code> should specify an initializer that contains
208    the final value to be read by jemalloc.  <code class="option">--with-malloc-conf</code>
209    and <code class="varname">malloc_conf</code> are compile-time mechanisms, whereas
210    <code class="filename">/etc/malloc.conf</code> and
211    <code class="envar">MALLOC_CONF</code> can be safely set any time prior to program
212    invocation.</p><p>An options string is a comma-separated list of option:value pairs.
213    There is one key corresponding to each <a class="link" href="#opt.abort"><quote><code class="mallctl">opt.*</code></quote></a> mallctl (see the <a class="xref" href="#mallctl_namespace" title="MALLCTL NAMESPACE">MALLCTL NAMESPACE</a> section for options
214    documentation).  For example, <code class="literal">abort:true,narenas:1</code> sets
215    the <a class="link" href="#opt.abort"><quote><code class="mallctl">opt.abort</code></quote></a> and <a class="link" href="#opt.narenas"><quote><code class="mallctl">opt.narenas</code></quote></a> options.  Some
216    options have boolean values (true/false), others have integer values (base
217    8, 10, or 16, depending on prefix), and yet others have raw string
218    values.</p></div><div class="refsect1"><a name="implementation_notes"/><h2>IMPLEMENTATION NOTES</h2><p>Traditionally, allocators have used
219    <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span> to obtain memory, which is
220    suboptimal for several reasons, including race conditions, increased
221    fragmentation, and artificial limitations on maximum usable memory.  If
222    <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span> is supported by the operating
223    system, this allocator uses both
224    <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> and
225    <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>, in that order of preference;
226    otherwise only <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> is used.</p><p>This allocator uses multiple arenas in order to reduce lock
227    contention for threaded programs on multi-processor systems.  This works
228    well with regard to threading scalability, but incurs some costs.  There is
229    a small fixed per-arena overhead, and additionally, arenas manage memory
230    completely independently of each other, which means a small fixed increase
231    in overall memory fragmentation.  These overheads are not generally an
232    issue, given the number of arenas normally used.  Note that using
233    substantially more arenas than the default is not likely to improve
234    performance, mainly due to reduced cache performance.  However, it may make
235    sense to reduce the number of arenas if an application does not make much
236    use of the allocation functions.</p><p>In addition to multiple arenas, this allocator supports
237    thread-specific caching, in order to make it possible to completely avoid
238    synchronization for most allocation requests.  Such caching allows very fast
239    allocation in the common case, but it increases memory usage and
240    fragmentation, since a bounded number of objects can remain allocated in
241    each thread cache.</p><p>Memory is conceptually broken into extents.  Extents are always
242    aligned to multiples of the page size.  This alignment makes it possible to
243    find metadata for user objects quickly.  User objects are broken into two
244    categories according to size: small and large.  Contiguous small objects
245    comprise a slab, which resides within a single extent, whereas large objects
246    each have their own extents backing them.</p><p>Small objects are managed in groups by slabs.  Each slab maintains
247    a bitmap to track which regions are in use.  Allocation requests that are no
248    more than half the quantum (8 or 16, depending on architecture) are rounded
249    up to the nearest power of two that is at least <code class="code">sizeof(<span class="type">double</span>)</code>.  All other object size
250    classes are multiples of the quantum, spaced such that there are four size
251    classes for each doubling in size, which limits internal fragmentation to
252    approximately 20% for all but the smallest size classes.  Small size classes
253    are smaller than four times the page size, and large size classes extend
254    from four times the page size up to the largest size class that does not
255    exceed <code class="constant">PTRDIFF_MAX</code>.</p><p>Allocations are packed tightly together, which can be an issue for
256    multi-threaded applications.  If you need to assure that allocations do not
257    suffer from cacheline sharing, round your allocation requests up to the
258    nearest multiple of the cacheline size, or specify cacheline alignment when
259    allocating.</p><p>The <code class="function">realloc()</code>,
260    <code class="function">rallocx()</code>, and
261    <code class="function">xallocx()</code> functions may resize allocations
262    without moving them under limited circumstances.  Unlike the
263    <code class="function">*allocx()</code> API, the standard API does not
264    officially round up the usable size of an allocation to the nearest size
265    class, so technically it is necessary to call
266    <code class="function">realloc()</code> to grow e.g. a 9-byte allocation to
267    16 bytes, or shrink a 16-byte allocation to 9 bytes.  Growth and shrinkage
268    trivially succeeds in place as long as the pre-size and post-size both round
269    up to the same size class.  No other API guarantees are made regarding
270    in-place resizing, but the current implementation also tries to resize large
271    allocations in place, as long as the pre-size and post-size are both large.
272    For shrinkage to succeed, the extent allocator must support splitting (see
273    <a class="link" href="#arena.i.extent_hooks"><quote><code class="mallctl">arena.&lt;i&gt;.extent_hooks</code></quote></a>).
274    Growth only succeeds if the trailing memory is currently available, and the
275    extent allocator supports merging.</p><p>Assuming 4 KiB pages and a 16-byte quantum on a 64-bit system, the
276    size classes in each category are as shown in <a class="xref" href="#size_classes" title="Table 1. Size classes">Table 1</a>.</p><div class="table"><a name="size_classes"/><p class="title"><b>Table 1. Size classes</b></p><div class="table-contents"><table summary="Size classes" border="1"><colgroup><col align="left" class="c1"/><col align="right" class="c2"/><col align="left" class="c3"/></colgroup><thead><tr><th align="left">Category</th><th align="right">Spacing</th><th align="left">Size</th></tr></thead><tbody><tr><td rowspan="9" align="left">Small</td><td align="right">lg</td><td align="left">[8]</td></tr><tr><td align="right">16</td><td align="left">[16, 32, 48, 64, 80, 96, 112, 128]</td></tr><tr><td align="right">32</td><td align="left">[160, 192, 224, 256]</td></tr><tr><td align="right">64</td><td align="left">[320, 384, 448, 512]</td></tr><tr><td align="right">128</td><td align="left">[640, 768, 896, 1024]</td></tr><tr><td align="right">256</td><td align="left">[1280, 1536, 1792, 2048]</td></tr><tr><td align="right">512</td><td align="left">[2560, 3072, 3584, 4096]</td></tr><tr><td align="right">1 KiB</td><td align="left">[5 KiB, 6 KiB, 7 KiB, 8 KiB]</td></tr><tr><td align="right">2 KiB</td><td align="left">[10 KiB, 12 KiB, 14 KiB]</td></tr><tr><td rowspan="16" align="left">Large</td><td align="right">2 KiB</td><td align="left">[16 KiB]</td></tr><tr><td align="right">4 KiB</td><td align="left">[20 KiB, 24 KiB, 28 KiB, 32 KiB]</td></tr><tr><td align="right">8 KiB</td><td align="left">[40 KiB, 48 KiB, 54 KiB, 64 KiB]</td></tr><tr><td align="right">16 KiB</td><td align="left">[80 KiB, 96 KiB, 112 KiB, 128 KiB]</td></tr><tr><td align="right">32 KiB</td><td align="left">[160 KiB, 192 KiB, 224 KiB, 256 KiB]</td></tr><tr><td align="right">64 KiB</td><td align="left">[320 KiB, 384 KiB, 448 KiB, 512 KiB]</td></tr><tr><td align="right">128 KiB</td><td align="left">[640 KiB, 768 KiB, 896 KiB, 1 MiB]</td></tr><tr><td align="right">256 KiB</td><td align="left">[1280 KiB, 1536 KiB, 1792 KiB, 2 MiB]</td></tr><tr><td align="right">512 KiB</td><td align="left">[2560 KiB, 3 MiB, 3584 KiB, 4 MiB]</td></tr><tr><td align="right">1 MiB</td><td align="left">[5 MiB, 6 MiB, 7 MiB, 8 MiB]</td></tr><tr><td align="right">2 MiB</td><td align="left">[10 MiB, 12 MiB, 14 MiB, 16 MiB]</td></tr><tr><td align="right">4 MiB</td><td align="left">[20 MiB, 24 MiB, 28 MiB, 32 MiB]</td></tr><tr><td align="right">8 MiB</td><td align="left">[40 MiB, 48 MiB, 56 MiB, 64 MiB]</td></tr><tr><td align="right">...</td><td align="left">...</td></tr><tr><td align="right">512 PiB</td><td align="left">[2560 PiB, 3 EiB, 3584 PiB, 4 EiB]</td></tr><tr><td align="right">1 EiB</td><td align="left">[5 EiB, 6 EiB, 7 EiB]</td></tr></tbody></table></div></div><br class="table-break"/></div><div class="refsect1"><a name="mallctl_namespace"/><h2>MALLCTL NAMESPACE</h2><p>The following names are defined in the namespace accessible via the
277    <code class="function">mallctl*()</code> functions.  Value types are specified in
278    parentheses, their readable/writable statuses are encoded as
279    <code class="literal">rw</code>, <code class="literal">r-</code>, <code class="literal">-w</code>, or
280    <code class="literal">--</code>, and required build configuration flags follow, if
281    any.  A name element encoded as <code class="literal">&lt;i&gt;</code> or
282    <code class="literal">&lt;j&gt;</code> indicates an integer component, where the
283    integer varies from 0 to some upper value that must be determined via
284    introspection.  In the case of <quote><code class="mallctl">stats.arenas.&lt;i&gt;.*</code></quote>
285    and <quote><code class="mallctl">arena.&lt;i&gt;.{initialized,purge,decay,dss}</code></quote>,
286    <code class="literal">&lt;i&gt;</code> equal to
287    <code class="constant">MALLCTL_ARENAS_ALL</code> can be used to operate on all arenas
288    or access the summation of statistics from all arenas; similarly
289    <code class="literal">&lt;i&gt;</code> equal to
290    <code class="constant">MALLCTL_ARENAS_DESTROYED</code> can be used to access the
291    summation of statistics from all destroyed arenas.  These constants can be
292    utilized either via <code class="function">mallctlnametomib()</code> followed by
293    <code class="function">mallctlbymib()</code>, or via code such as the following:
294    </p><pre class="programlisting">
295#define STRINGIFY_HELPER(x) #x
296#define STRINGIFY(x) STRINGIFY_HELPER(x)
297
298mallctl("arena." STRINGIFY(MALLCTL_ARENAS_ALL) ".decay",
299    NULL, NULL, NULL, 0);</pre><p>
300    Take special note of the <a class="link" href="#epoch"><quote><code class="mallctl">epoch</code></quote></a> mallctl, which controls
301    refreshing of cached dynamic statistics.</p><div class="variablelist"><dl class="variablelist"><dt><a name="version"/><span class="term">
302          <quote><code class="mallctl">version</code></quote>
303          (<span class="type">const char *</span>)
304          <code class="literal">r-</code>
305        </span></dt><dd><p>Return the jemalloc version string.</p></dd><dt><a name="epoch"/><span class="term">
306          <quote><code class="mallctl">epoch</code></quote>
307          (<span class="type">uint64_t</span>)
308          <code class="literal">rw</code>
309        </span></dt><dd><p>If a value is passed in, refresh the data from which
310        the <code class="function">mallctl*()</code> functions report values,
311        and increment the epoch.  Return the current epoch.  This is useful for
312        detecting whether another thread caused a refresh.</p></dd><dt><a name="background_thread"/><span class="term">
313          <quote><code class="mallctl">background_thread</code></quote>
314          (<span class="type">bool</span>)
315          <code class="literal">rw</code>
316        </span></dt><dd><p>Enable/disable internal background worker threads.  When
317        set to true, background threads are created on demand (the number of
318        background threads will be no more than the number of CPUs or active
319        arenas).  Threads run periodically, and handle <a class="link" href="#arena.i.decay">purging</a> asynchronously.  When switching
320        off, background threads are terminated synchronously.  Note that after
321        <span class="citerefentry"><span class="refentrytitle">fork</span>(2)</span>
322        function, the state in the child process will be disabled regardless
323        the state in parent process. See <a class="link" href="#stats.background_thread.num_threads"><quote><code class="mallctl">stats.background_thread</code></quote></a>
324        for related stats.  <a class="link" href="#opt.background_thread"><quote><code class="mallctl">opt.background_thread</code></quote></a>
325        can be used to set the default option.  This option is only available on
326        selected pthread-based platforms.</p></dd><dt><a name="max_background_threads"/><span class="term">
327          <quote><code class="mallctl">max_background_threads</code></quote>
328          (<span class="type">size_t</span>)
329          <code class="literal">rw</code>
330        </span></dt><dd><p>Maximum number of background worker threads that will
331        be created.  This value is capped at <a class="link" href="#opt.max_background_threads"><quote><code class="mallctl">opt.max_background_threads</code></quote></a> at
332        startup.</p></dd><dt><a name="config.cache_oblivious"/><span class="term">
333          <quote><code class="mallctl">config.cache_oblivious</code></quote>
334          (<span class="type">bool</span>)
335          <code class="literal">r-</code>
336        </span></dt><dd><p><code class="option">--enable-cache-oblivious</code> was specified
337        during build configuration.</p></dd><dt><a name="config.debug"/><span class="term">
338          <quote><code class="mallctl">config.debug</code></quote>
339          (<span class="type">bool</span>)
340          <code class="literal">r-</code>
341        </span></dt><dd><p><code class="option">--enable-debug</code> was specified during
342        build configuration.</p></dd><dt><a name="config.fill"/><span class="term">
343          <quote><code class="mallctl">config.fill</code></quote>
344          (<span class="type">bool</span>)
345          <code class="literal">r-</code>
346        </span></dt><dd><p><code class="option">--enable-fill</code> was specified during
347        build configuration.</p></dd><dt><a name="config.lazy_lock"/><span class="term">
348          <quote><code class="mallctl">config.lazy_lock</code></quote>
349          (<span class="type">bool</span>)
350          <code class="literal">r-</code>
351        </span></dt><dd><p><code class="option">--enable-lazy-lock</code> was specified
352        during build configuration.</p></dd><dt><a name="config.malloc_conf"/><span class="term">
353          <quote><code class="mallctl">config.malloc_conf</code></quote>
354          (<span class="type">const char *</span>)
355          <code class="literal">r-</code>
356        </span></dt><dd><p>Embedded configure-time-specified run-time options
357        string, empty unless <code class="option">--with-malloc-conf</code> was specified
358        during build configuration.</p></dd><dt><a name="config.prof"/><span class="term">
359          <quote><code class="mallctl">config.prof</code></quote>
360          (<span class="type">bool</span>)
361          <code class="literal">r-</code>
362        </span></dt><dd><p><code class="option">--enable-prof</code> was specified during
363        build configuration.</p></dd><dt><a name="config.prof_libgcc"/><span class="term">
364          <quote><code class="mallctl">config.prof_libgcc</code></quote>
365          (<span class="type">bool</span>)
366          <code class="literal">r-</code>
367        </span></dt><dd><p><code class="option">--disable-prof-libgcc</code> was not
368        specified during build configuration.</p></dd><dt><a name="config.prof_libunwind"/><span class="term">
369          <quote><code class="mallctl">config.prof_libunwind</code></quote>
370          (<span class="type">bool</span>)
371          <code class="literal">r-</code>
372        </span></dt><dd><p><code class="option">--enable-prof-libunwind</code> was specified
373        during build configuration.</p></dd><dt><a name="config.stats"/><span class="term">
374          <quote><code class="mallctl">config.stats</code></quote>
375          (<span class="type">bool</span>)
376          <code class="literal">r-</code>
377        </span></dt><dd><p><code class="option">--enable-stats</code> was specified during
378        build configuration.</p></dd><dt><a name="config.utrace"/><span class="term">
379          <quote><code class="mallctl">config.utrace</code></quote>
380          (<span class="type">bool</span>)
381          <code class="literal">r-</code>
382        </span></dt><dd><p><code class="option">--enable-utrace</code> was specified during
383        build configuration.</p></dd><dt><a name="config.xmalloc"/><span class="term">
384          <quote><code class="mallctl">config.xmalloc</code></quote>
385          (<span class="type">bool</span>)
386          <code class="literal">r-</code>
387        </span></dt><dd><p><code class="option">--enable-xmalloc</code> was specified during
388        build configuration.</p></dd><dt><a name="opt.abort"/><span class="term">
389          <quote><code class="mallctl">opt.abort</code></quote>
390          (<span class="type">bool</span>)
391          <code class="literal">r-</code>
392        </span></dt><dd><p>Abort-on-warning enabled/disabled.  If true, most
393        warnings are fatal.  Note that runtime option warnings are not included
394        (see <a class="link" href="#opt.abort_conf"><quote><code class="mallctl">opt.abort_conf</code></quote></a> for
395        that). The process will call
396        <span class="citerefentry"><span class="refentrytitle">abort</span>(3)</span> in these cases.  This option is
397        disabled by default unless <code class="option">--enable-debug</code> is
398        specified during configuration, in which case it is enabled by default.
399        </p></dd><dt><a name="opt.abort_conf"/><span class="term">
400          <quote><code class="mallctl">opt.abort_conf</code></quote>
401          (<span class="type">bool</span>)
402          <code class="literal">r-</code>
403        </span></dt><dd><p>Abort-on-invalid-configuration enabled/disabled.  If
404        true, invalid runtime options are fatal.  The process will call
405        <span class="citerefentry"><span class="refentrytitle">abort</span>(3)</span> in these cases.  This option is
406        disabled by default unless <code class="option">--enable-debug</code> is
407        specified during configuration, in which case it is enabled by default.
408        </p></dd><dt><a name="opt.metadata_thp"/><span class="term">
409          <quote><code class="mallctl">opt.metadata_thp</code></quote>
410          (<span class="type">const char *</span>)
411          <code class="literal">r-</code>
412        </span></dt><dd><p>Controls whether to allow jemalloc to use transparent
413        huge page (THP) for internal metadata (see <a class="link" href="#stats.metadata">stats.metadata</a>).  <span class="quote">“<span class="quote">always</span>”</span>
414        allows such usage.  <span class="quote">“<span class="quote">auto</span>”</span> uses no THP initially, but may
415        begin to do so when metadata usage reaches certain level.  The default
416        is <span class="quote">“<span class="quote">disabled</span>”</span>.</p></dd><dt><a name="opt.retain"/><span class="term">
417          <quote><code class="mallctl">opt.retain</code></quote>
418          (<span class="type">bool</span>)
419          <code class="literal">r-</code>
420        </span></dt><dd><p>If true, retain unused virtual memory for later reuse
421        rather than discarding it by calling
422        <span class="citerefentry"><span class="refentrytitle">munmap</span>(2)</span> or equivalent (see <a class="link" href="#stats.retained">stats.retained</a> for related details).
423        This option is disabled by default unless discarding virtual memory is
424        known to trigger
425        platform-specific performance problems, e.g. for [64-bit] Linux, which
426        has a quirk in its virtual memory allocation algorithm that causes
427        semi-permanent VM map holes under normal jemalloc operation.  Although
428        <span class="citerefentry"><span class="refentrytitle">munmap</span>(2)</span> causes issues on 32-bit Linux as
429        well, retaining virtual memory for 32-bit Linux is disabled by default
430        due to the practical possibility of address space exhaustion.
431        </p></dd><dt><a name="opt.dss"/><span class="term">
432          <quote><code class="mallctl">opt.dss</code></quote>
433          (<span class="type">const char *</span>)
434          <code class="literal">r-</code>
435        </span></dt><dd><p>dss (<span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>) allocation precedence as
436        related to <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> allocation.  The following
437        settings are supported if
438        <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span> is supported by the operating
439        system: <span class="quote">“<span class="quote">disabled</span>”</span>, <span class="quote">“<span class="quote">primary</span>”</span>, and
440        <span class="quote">“<span class="quote">secondary</span>”</span>; otherwise only <span class="quote">“<span class="quote">disabled</span>”</span> is
441        supported.  The default is <span class="quote">“<span class="quote">secondary</span>”</span> if
442        <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span> is supported by the operating
443        system; <span class="quote">“<span class="quote">disabled</span>”</span> otherwise.
444        </p></dd><dt><a name="opt.narenas"/><span class="term">
445          <quote><code class="mallctl">opt.narenas</code></quote>
446          (<span class="type">unsigned</span>)
447          <code class="literal">r-</code>
448        </span></dt><dd><p>Maximum number of arenas to use for automatic
449        multiplexing of threads and arenas.  The default is four times the
450        number of CPUs, or one if there is a single CPU.</p></dd><dt><a name="opt.percpu_arena"/><span class="term">
451          <quote><code class="mallctl">opt.percpu_arena</code></quote>
452          (<span class="type">const char *</span>)
453          <code class="literal">r-</code>
454        </span></dt><dd><p>Per CPU arena mode.  Use the <span class="quote">“<span class="quote">percpu</span>”</span>
455        setting to enable this feature, which uses number of CPUs to determine
456        number of arenas, and bind threads to arenas dynamically based on the
457        CPU the thread runs on currently.  <span class="quote">“<span class="quote">phycpu</span>”</span> setting uses
458        one arena per physical CPU, which means the two hyper threads on the
459        same CPU share one arena.  Note that no runtime checking regarding the
460        availability of hyper threading is done at the moment.  When set to
461        <span class="quote">“<span class="quote">disabled</span>”</span>, narenas and thread to arena association will
462        not be impacted by this option.  The default is <span class="quote">“<span class="quote">disabled</span>”</span>.
463        </p></dd><dt><a name="opt.background_thread"/><span class="term">
464          <quote><code class="mallctl">opt.background_thread</code></quote>
465          (<span class="type">const bool</span>)
466          <code class="literal">r-</code>
467        </span></dt><dd><p>Internal background worker threads enabled/disabled.
468        Because of potential circular dependencies, enabling background thread
469        using this option may cause crash or deadlock during initialization. For
470        a reliable way to use this feature, see <a class="link" href="#background_thread">background_thread</a> for dynamic control
471        options and details.  This option is disabled by
472        default.</p></dd><dt><a name="opt.max_background_threads"/><span class="term">
473          <quote><code class="mallctl">opt.max_background_threads</code></quote>
474          (<span class="type">const size_t</span>)
475          <code class="literal">r-</code>
476        </span></dt><dd><p>Maximum number of background threads that will be created
477        if <a class="link" href="#background_thread">background_thread</a> is set.
478        Defaults to number of cpus.</p></dd><dt><a name="opt.dirty_decay_ms"/><span class="term">
479          <quote><code class="mallctl">opt.dirty_decay_ms</code></quote>
480          (<span class="type">ssize_t</span>)
481          <code class="literal">r-</code>
482        </span></dt><dd><p>Approximate time in milliseconds from the creation of a
483        set of unused dirty pages until an equivalent set of unused dirty pages
484        is purged (i.e. converted to muzzy via e.g.
485        <code class="function">madvise(<em class="parameter"><code>...</code></em><em class="parameter"><code><code class="constant">MADV_FREE</code></code></em>)</code>
486        if supported by the operating system, or converted to clean otherwise)
487        and/or reused.  Dirty pages are defined as previously having been
488        potentially written to by the application, and therefore consuming
489        physical memory, yet having no current use.  The pages are incrementally
490        purged according to a sigmoidal decay curve that starts and ends with
491        zero purge rate.  A decay time of 0 causes all unused dirty pages to be
492        purged immediately upon creation.  A decay time of -1 disables purging.
493        The default decay time is 10 seconds.  See <a class="link" href="#arenas.dirty_decay_ms"><quote><code class="mallctl">arenas.dirty_decay_ms</code></quote></a>
494        and <a class="link" href="#arena.i.dirty_decay_ms"><quote><code class="mallctl">arena.&lt;i&gt;.dirty_decay_ms</code></quote></a>
495        for related dynamic control options.  See <a class="link" href="#opt.muzzy_decay_ms"><quote><code class="mallctl">opt.muzzy_decay_ms</code></quote></a>
496        for a description of muzzy pages.</p></dd><dt><a name="opt.muzzy_decay_ms"/><span class="term">
497          <quote><code class="mallctl">opt.muzzy_decay_ms</code></quote>
498          (<span class="type">ssize_t</span>)
499          <code class="literal">r-</code>
500        </span></dt><dd><p>Approximate time in milliseconds from the creation of a
501        set of unused muzzy pages until an equivalent set of unused muzzy pages
502        is purged (i.e. converted to clean) and/or reused.  Muzzy pages are
503        defined as previously having been unused dirty pages that were
504        subsequently purged in a manner that left them subject to the
505        reclamation whims of the operating system (e.g.
506        <code class="function">madvise(<em class="parameter"><code>...</code></em><em class="parameter"><code><code class="constant">MADV_FREE</code></code></em>)</code>),
507        and therefore in an indeterminate state.  The pages are incrementally
508        purged according to a sigmoidal decay curve that starts and ends with
509        zero purge rate.  A decay time of 0 causes all unused muzzy pages to be
510        purged immediately upon creation.  A decay time of -1 disables purging.
511        The default decay time is 10 seconds.  See <a class="link" href="#arenas.muzzy_decay_ms"><quote><code class="mallctl">arenas.muzzy_decay_ms</code></quote></a>
512        and <a class="link" href="#arena.i.muzzy_decay_ms"><quote><code class="mallctl">arena.&lt;i&gt;.muzzy_decay_ms</code></quote></a>
513        for related dynamic control options.</p></dd><dt><a name="opt.lg_extent_max_active_fit"/><span class="term">
514          <quote><code class="mallctl">opt.lg_extent_max_active_fit</code></quote>
515          (<span class="type">size_t</span>)
516          <code class="literal">r-</code>
517        </span></dt><dd><p>When reusing dirty extents, this determines the (log
518        base 2 of the) maximum ratio between the size of the active extent
519        selected (to split off from) and the size of the requested allocation.
520        This prevents the splitting of large active extents for smaller
521        allocations, which can reduce fragmentation over the long run
522        (especially for non-active extents).  Lower value may reduce
523        fragmentation, at the cost of extra active extents.  The default value
524        is 6, which gives a maximum ratio of 64 (2^6).</p></dd><dt><a name="opt.stats_print"/><span class="term">
525          <quote><code class="mallctl">opt.stats_print</code></quote>
526          (<span class="type">bool</span>)
527          <code class="literal">r-</code>
528        </span></dt><dd><p>Enable/disable statistics printing at exit.  If
529        enabled, the <code class="function">malloc_stats_print()</code>
530        function is called at program exit via an
531        <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function.  <a class="link" href="#opt.stats_print_opts"><quote><code class="mallctl">opt.stats_print_opts</code></quote></a>
532        can be combined to specify output options. If
533        <code class="option">--enable-stats</code> is specified during configuration, this
534        has the potential to cause deadlock for a multi-threaded process that
535        exits while one or more threads are executing in the memory allocation
536        functions.  Furthermore, <code class="function">atexit()</code> may
537        allocate memory during application initialization and then deadlock
538        internally when jemalloc in turn calls
539        <code class="function">atexit()</code>, so this option is not
540        universally usable (though the application can register its own
541        <code class="function">atexit()</code> function with equivalent
542        functionality).  Therefore, this option should only be used with care;
543        it is primarily intended as a performance tuning aid during application
544        development.  This option is disabled by default.</p></dd><dt><a name="opt.stats_print_opts"/><span class="term">
545          <quote><code class="mallctl">opt.stats_print_opts</code></quote>
546          (<span class="type">const char *</span>)
547          <code class="literal">r-</code>
548        </span></dt><dd><p>Options (the <em class="parameter"><code>opts</code></em> string) to pass
549        to the <code class="function">malloc_stats_print()</code> at exit (enabled
550        through <a class="link" href="#opt.stats_print"><quote><code class="mallctl">opt.stats_print</code></quote></a>). See
551        available options in <a class="link" href="#malloc_stats_print_opts"><code class="function">malloc_stats_print()</code></a>.
552        Has no effect unless <a class="link" href="#opt.stats_print"><quote><code class="mallctl">opt.stats_print</code></quote></a> is
553        enabled.  The default is <span class="quote">“<span class="quote"/>”</span>.</p></dd><dt><a name="opt.junk"/><span class="term">
554          <quote><code class="mallctl">opt.junk</code></quote>
555          (<span class="type">const char *</span>)
556          <code class="literal">r-</code>
557          [<code class="option">--enable-fill</code>]
558        </span></dt><dd><p>Junk filling.  If set to <span class="quote">“<span class="quote">alloc</span>”</span>, each byte
559        of uninitialized allocated memory will be initialized to
560        <code class="literal">0xa5</code>.  If set to <span class="quote">“<span class="quote">free</span>”</span>, all deallocated
561        memory will be initialized to <code class="literal">0x5a</code>.  If set to
562        <span class="quote">“<span class="quote">true</span>”</span>, both allocated and deallocated memory will be
563        initialized, and if set to <span class="quote">“<span class="quote">false</span>”</span>, junk filling be
564        disabled entirely.  This is intended for debugging and will impact
565        performance negatively.  This option is <span class="quote">“<span class="quote">false</span>”</span> by default
566        unless <code class="option">--enable-debug</code> is specified during
567        configuration, in which case it is <span class="quote">“<span class="quote">true</span>”</span> by
568        default.</p></dd><dt><a name="opt.zero"/><span class="term">
569          <quote><code class="mallctl">opt.zero</code></quote>
570          (<span class="type">bool</span>)
571          <code class="literal">r-</code>
572          [<code class="option">--enable-fill</code>]
573        </span></dt><dd><p>Zero filling enabled/disabled.  If enabled, each byte
574        of uninitialized allocated memory will be initialized to 0.  Note that
575        this initialization only happens once for each byte, so
576        <code class="function">realloc()</code> and
577        <code class="function">rallocx()</code> calls do not zero memory that
578        was previously allocated.  This is intended for debugging and will
579        impact performance negatively.  This option is disabled by default.
580        </p></dd><dt><a name="opt.utrace"/><span class="term">
581          <quote><code class="mallctl">opt.utrace</code></quote>
582          (<span class="type">bool</span>)
583          <code class="literal">r-</code>
584          [<code class="option">--enable-utrace</code>]
585        </span></dt><dd><p>Allocation tracing based on
586        <span class="citerefentry"><span class="refentrytitle">utrace</span>(2)</span> enabled/disabled.  This option
587        is disabled by default.</p></dd><dt><a name="opt.xmalloc"/><span class="term">
588          <quote><code class="mallctl">opt.xmalloc</code></quote>
589          (<span class="type">bool</span>)
590          <code class="literal">r-</code>
591          [<code class="option">--enable-xmalloc</code>]
592        </span></dt><dd><p>Abort-on-out-of-memory enabled/disabled.  If enabled,
593        rather than returning failure for any allocation function, display a
594        diagnostic message on <code class="constant">STDERR_FILENO</code> and cause the
595        program to drop core (using
596        <span class="citerefentry"><span class="refentrytitle">abort</span>(3)</span>).  If an application is
597        designed to depend on this behavior, set the option at compile time by
598        including the following in the source code:
599        </p><pre class="programlisting">
600malloc_conf = "xmalloc:true";</pre><p>
601        This option is disabled by default.</p></dd><dt><a name="opt.tcache"/><span class="term">
602          <quote><code class="mallctl">opt.tcache</code></quote>
603          (<span class="type">bool</span>)
604          <code class="literal">r-</code>
605        </span></dt><dd><p>Thread-specific caching (tcache) enabled/disabled.  When
606        there are multiple threads, each thread uses a tcache for objects up to
607        a certain size.  Thread-specific caching allows many allocations to be
608        satisfied without performing any thread synchronization, at the cost of
609        increased memory use.  See the <a class="link" href="#opt.lg_tcache_max"><quote><code class="mallctl">opt.lg_tcache_max</code></quote></a>
610        option for related tuning information.  This option is enabled by
611        default.</p></dd><dt><a name="opt.lg_tcache_max"/><span class="term">
612          <quote><code class="mallctl">opt.lg_tcache_max</code></quote>
613          (<span class="type">size_t</span>)
614          <code class="literal">r-</code>
615        </span></dt><dd><p>Maximum size class (log base 2) to cache in the
616        thread-specific cache (tcache).  At a minimum, all small size classes
617        are cached, and at a maximum all large size classes are cached.  The
618        default maximum is 32 KiB (2^15).</p></dd><dt><a name="opt.thp"/><span class="term">
619          <quote><code class="mallctl">opt.thp</code></quote>
620          (<span class="type">const char *</span>)
621          <code class="literal">r-</code>
622        </span></dt><dd><p>Transparent hugepage (THP) mode. Settings "always",
623        "never" and "default" are available if THP is supported by the operating
624        system.  The "always" setting enables transparent hugepage for all user
625        memory mappings with
626        <em class="parameter"><code><code class="constant">MADV_HUGEPAGE</code></code></em>; "never"
627        ensures no transparent hugepage with
628        <em class="parameter"><code><code class="constant">MADV_NOHUGEPAGE</code></code></em>; the default
629        setting "default" makes no changes.  Note that: this option does not
630        affect THP for jemalloc internal metadata (see <a class="link" href="#opt.metadata_thp"><quote><code class="mallctl">opt.metadata_thp</code></quote></a>);
631        in addition, for arenas with customized <a class="link" href="#arena.i.extent_hooks"><quote><code class="mallctl">extent_hooks</code></quote></a>,
632        this option is bypassed as it is implemented as part of the default
633        extent hooks.</p></dd><dt><a name="opt.prof"/><span class="term">
634          <quote><code class="mallctl">opt.prof</code></quote>
635          (<span class="type">bool</span>)
636          <code class="literal">r-</code>
637          [<code class="option">--enable-prof</code>]
638        </span></dt><dd><p>Memory profiling enabled/disabled.  If enabled, profile
639        memory allocation activity.  See the <a class="link" href="#opt.prof_active"><quote><code class="mallctl">opt.prof_active</code></quote></a>
640        option for on-the-fly activation/deactivation.  See the <a class="link" href="#opt.lg_prof_sample"><quote><code class="mallctl">opt.lg_prof_sample</code></quote></a>
641        option for probabilistic sampling control.  See the <a class="link" href="#opt.prof_accum"><quote><code class="mallctl">opt.prof_accum</code></quote></a>
642        option for control of cumulative sample reporting.  See the <a class="link" href="#opt.lg_prof_interval"><quote><code class="mallctl">opt.lg_prof_interval</code></quote></a>
643        option for information on interval-triggered profile dumping, the <a class="link" href="#opt.prof_gdump"><quote><code class="mallctl">opt.prof_gdump</code></quote></a>
644        option for information on high-water-triggered profile dumping, and the
645        <a class="link" href="#opt.prof_final"><quote><code class="mallctl">opt.prof_final</code></quote></a>
646        option for final profile dumping.  Profile output is compatible with
647        the <span class="command"><strong>jeprof</strong></span> command, which is based on the
648        <span class="command"><strong>pprof</strong></span> that is developed as part of the <a class="ulink" href="http://code.google.com/p/gperftools/" target="_top">gperftools
649        package</a>.  See <a class="link" href="#heap_profile_format" title="HEAP PROFILE FORMAT">HEAP PROFILE
650        FORMAT</a> for heap profile format documentation.</p></dd><dt><a name="opt.prof_prefix"/><span class="term">
651          <quote><code class="mallctl">opt.prof_prefix</code></quote>
652          (<span class="type">const char *</span>)
653          <code class="literal">r-</code>
654          [<code class="option">--enable-prof</code>]
655        </span></dt><dd><p>Filename prefix for profile dumps.  If the prefix is
656        set to the empty string, no automatic dumps will occur; this is
657        primarily useful for disabling the automatic final heap dump (which
658        also disables leak reporting, if enabled).  The default prefix is
659        <code class="filename">jeprof</code>.</p></dd><dt><a name="opt.prof_active"/><span class="term">
660          <quote><code class="mallctl">opt.prof_active</code></quote>
661          (<span class="type">bool</span>)
662          <code class="literal">r-</code>
663          [<code class="option">--enable-prof</code>]
664        </span></dt><dd><p>Profiling activated/deactivated.  This is a secondary
665        control mechanism that makes it possible to start the application with
666        profiling enabled (see the <a class="link" href="#opt.prof"><quote><code class="mallctl">opt.prof</code></quote></a> option) but
667        inactive, then toggle profiling at any time during program execution
668        with the <a class="link" href="#prof.active"><quote><code class="mallctl">prof.active</code></quote></a> mallctl.
669        This option is enabled by default.</p></dd><dt><a name="opt.prof_thread_active_init"/><span class="term">
670          <quote><code class="mallctl">opt.prof_thread_active_init</code></quote>
671          (<span class="type">bool</span>)
672          <code class="literal">r-</code>
673          [<code class="option">--enable-prof</code>]
674        </span></dt><dd><p>Initial setting for <a class="link" href="#thread.prof.active"><quote><code class="mallctl">thread.prof.active</code></quote></a>
675        in newly created threads.  The initial setting for newly created threads
676        can also be changed during execution via the <a class="link" href="#prof.thread_active_init"><quote><code class="mallctl">prof.thread_active_init</code></quote></a>
677        mallctl.  This option is enabled by default.</p></dd><dt><a name="opt.lg_prof_sample"/><span class="term">
678          <quote><code class="mallctl">opt.lg_prof_sample</code></quote>
679          (<span class="type">size_t</span>)
680          <code class="literal">r-</code>
681          [<code class="option">--enable-prof</code>]
682        </span></dt><dd><p>Average interval (log base 2) between allocation
683        samples, as measured in bytes of allocation activity.  Increasing the
684        sampling interval decreases profile fidelity, but also decreases the
685        computational overhead.  The default sample interval is 512 KiB (2^19
686        B).</p></dd><dt><a name="opt.prof_accum"/><span class="term">
687          <quote><code class="mallctl">opt.prof_accum</code></quote>
688          (<span class="type">bool</span>)
689          <code class="literal">r-</code>
690          [<code class="option">--enable-prof</code>]
691        </span></dt><dd><p>Reporting of cumulative object/byte counts in profile
692        dumps enabled/disabled.  If this option is enabled, every unique
693        backtrace must be stored for the duration of execution.  Depending on
694        the application, this can impose a large memory overhead, and the
695        cumulative counts are not always of interest.  This option is disabled
696        by default.</p></dd><dt><a name="opt.lg_prof_interval"/><span class="term">
697          <quote><code class="mallctl">opt.lg_prof_interval</code></quote>
698          (<span class="type">ssize_t</span>)
699          <code class="literal">r-</code>
700          [<code class="option">--enable-prof</code>]
701        </span></dt><dd><p>Average interval (log base 2) between memory profile
702        dumps, as measured in bytes of allocation activity.  The actual
703        interval between dumps may be sporadic because decentralized allocation
704        counters are used to avoid synchronization bottlenecks.  Profiles are
705        dumped to files named according to the pattern
706        <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.i&lt;iseq&gt;.heap</code>,
707        where <code class="literal">&lt;prefix&gt;</code> is controlled by the
708        <a class="link" href="#opt.prof_prefix"><quote><code class="mallctl">opt.prof_prefix</code></quote></a>
709        option.  By default, interval-triggered profile dumping is disabled
710        (encoded as -1).
711        </p></dd><dt><a name="opt.prof_gdump"/><span class="term">
712          <quote><code class="mallctl">opt.prof_gdump</code></quote>
713          (<span class="type">bool</span>)
714          <code class="literal">r-</code>
715          [<code class="option">--enable-prof</code>]
716        </span></dt><dd><p>Set the initial state of <a class="link" href="#prof.gdump"><quote><code class="mallctl">prof.gdump</code></quote></a>, which when
717        enabled triggers a memory profile dump every time the total virtual
718        memory exceeds the previous maximum.  This option is disabled by
719        default.</p></dd><dt><a name="opt.prof_final"/><span class="term">
720          <quote><code class="mallctl">opt.prof_final</code></quote>
721          (<span class="type">bool</span>)
722          <code class="literal">r-</code>
723          [<code class="option">--enable-prof</code>]
724        </span></dt><dd><p>Use an
725        <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function to dump final memory
726        usage to a file named according to the pattern
727        <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.f.heap</code>,
728        where <code class="literal">&lt;prefix&gt;</code> is controlled by the <a class="link" href="#opt.prof_prefix"><quote><code class="mallctl">opt.prof_prefix</code></quote></a>
729        option.  Note that <code class="function">atexit()</code> may allocate
730        memory during application initialization and then deadlock internally
731        when jemalloc in turn calls <code class="function">atexit()</code>, so
732        this option is not universally usable (though the application can
733        register its own <code class="function">atexit()</code> function with
734        equivalent functionality).  This option is disabled by
735        default.</p></dd><dt><a name="opt.prof_leak"/><span class="term">
736          <quote><code class="mallctl">opt.prof_leak</code></quote>
737          (<span class="type">bool</span>)
738          <code class="literal">r-</code>
739          [<code class="option">--enable-prof</code>]
740        </span></dt><dd><p>Leak reporting enabled/disabled.  If enabled, use an
741        <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function to report memory leaks
742        detected by allocation sampling.  See the
743        <a class="link" href="#opt.prof"><quote><code class="mallctl">opt.prof</code></quote></a> option for
744        information on analyzing heap profile output.  This option is disabled
745        by default.</p></dd><dt><a name="thread.arena"/><span class="term">
746          <quote><code class="mallctl">thread.arena</code></quote>
747          (<span class="type">unsigned</span>)
748          <code class="literal">rw</code>
749        </span></dt><dd><p>Get or set the arena associated with the calling
750        thread.  If the specified arena was not initialized beforehand (see the
751        <a class="link" href="#arena.i.initialized"><quote><code class="mallctl">arena.i.initialized</code></quote></a>
752        mallctl), it will be automatically initialized as a side effect of
753        calling this interface.</p></dd><dt><a name="thread.allocated"/><span class="term">
754          <quote><code class="mallctl">thread.allocated</code></quote>
755          (<span class="type">uint64_t</span>)
756          <code class="literal">r-</code>
757          [<code class="option">--enable-stats</code>]
758        </span></dt><dd><p>Get the total number of bytes ever allocated by the
759        calling thread.  This counter has the potential to wrap around; it is
760        up to the application to appropriately interpret the counter in such
761        cases.</p></dd><dt><a name="thread.allocatedp"/><span class="term">
762          <quote><code class="mallctl">thread.allocatedp</code></quote>
763          (<span class="type">uint64_t *</span>)
764          <code class="literal">r-</code>
765          [<code class="option">--enable-stats</code>]
766        </span></dt><dd><p>Get a pointer to the the value that is returned by the
767        <a class="link" href="#thread.allocated"><quote><code class="mallctl">thread.allocated</code></quote></a>
768        mallctl.  This is useful for avoiding the overhead of repeated
769        <code class="function">mallctl*()</code> calls.</p></dd><dt><a name="thread.deallocated"/><span class="term">
770          <quote><code class="mallctl">thread.deallocated</code></quote>
771          (<span class="type">uint64_t</span>)
772          <code class="literal">r-</code>
773          [<code class="option">--enable-stats</code>]
774        </span></dt><dd><p>Get the total number of bytes ever deallocated by the
775        calling thread.  This counter has the potential to wrap around; it is
776        up to the application to appropriately interpret the counter in such
777        cases.</p></dd><dt><a name="thread.deallocatedp"/><span class="term">
778          <quote><code class="mallctl">thread.deallocatedp</code></quote>
779          (<span class="type">uint64_t *</span>)
780          <code class="literal">r-</code>
781          [<code class="option">--enable-stats</code>]
782        </span></dt><dd><p>Get a pointer to the the value that is returned by the
783        <a class="link" href="#thread.deallocated"><quote><code class="mallctl">thread.deallocated</code></quote></a>
784        mallctl.  This is useful for avoiding the overhead of repeated
785        <code class="function">mallctl*()</code> calls.</p></dd><dt><a name="thread.tcache.enabled"/><span class="term">
786          <quote><code class="mallctl">thread.tcache.enabled</code></quote>
787          (<span class="type">bool</span>)
788          <code class="literal">rw</code>
789        </span></dt><dd><p>Enable/disable calling thread's tcache.  The tcache is
790        implicitly flushed as a side effect of becoming
791        disabled (see <a class="link" href="#thread.tcache.flush"><quote><code class="mallctl">thread.tcache.flush</code></quote></a>).
792        </p></dd><dt><a name="thread.tcache.flush"/><span class="term">
793          <quote><code class="mallctl">thread.tcache.flush</code></quote>
794          (<span class="type">void</span>)
795          <code class="literal">--</code>
796        </span></dt><dd><p>Flush calling thread's thread-specific cache (tcache).
797        This interface releases all cached objects and internal data structures
798        associated with the calling thread's tcache.  Ordinarily, this interface
799        need not be called, since automatic periodic incremental garbage
800        collection occurs, and the thread cache is automatically discarded when
801        a thread exits.  However, garbage collection is triggered by allocation
802        activity, so it is possible for a thread that stops
803        allocating/deallocating to retain its cache indefinitely, in which case
804        the developer may find manual flushing useful.</p></dd><dt><a name="thread.prof.name"/><span class="term">
805          <quote><code class="mallctl">thread.prof.name</code></quote>
806          (<span class="type">const char *</span>)
807          <code class="literal">r-</code> or
808          <code class="literal">-w</code>
809          [<code class="option">--enable-prof</code>]
810        </span></dt><dd><p>Get/set the descriptive name associated with the calling
811        thread in memory profile dumps.  An internal copy of the name string is
812        created, so the input string need not be maintained after this interface
813        completes execution.  The output string of this interface should be
814        copied for non-ephemeral uses, because multiple implementation details
815        can cause asynchronous string deallocation.  Furthermore, each
816        invocation of this interface can only read or write; simultaneous
817        read/write is not supported due to string lifetime limitations.  The
818        name string must be nil-terminated and comprised only of characters in
819        the sets recognized
820        by <span class="citerefentry"><span class="refentrytitle">isgraph</span>(3)</span> and
821        <span class="citerefentry"><span class="refentrytitle">isblank</span>(3)</span>.</p></dd><dt><a name="thread.prof.active"/><span class="term">
822          <quote><code class="mallctl">thread.prof.active</code></quote>
823          (<span class="type">bool</span>)
824          <code class="literal">rw</code>
825          [<code class="option">--enable-prof</code>]
826        </span></dt><dd><p>Control whether sampling is currently active for the
827        calling thread.  This is an activation mechanism in addition to <a class="link" href="#prof.active"><quote><code class="mallctl">prof.active</code></quote></a>; both must
828        be active for the calling thread to sample.  This flag is enabled by
829        default.</p></dd><dt><a name="tcache.create"/><span class="term">
830          <quote><code class="mallctl">tcache.create</code></quote>
831          (<span class="type">unsigned</span>)
832          <code class="literal">r-</code>
833        </span></dt><dd><p>Create an explicit thread-specific cache (tcache) and
834        return an identifier that can be passed to the <a class="link" href="#MALLOCX_TCACHE"><code class="constant">MALLOCX_TCACHE(<em class="parameter"><code>tc</code></em>)</code></a>
835        macro to explicitly use the specified cache rather than the
836        automatically managed one that is used by default.  Each explicit cache
837        can be used by only one thread at a time; the application must assure
838        that this constraint holds.
839        </p></dd><dt><a name="tcache.flush"/><span class="term">
840          <quote><code class="mallctl">tcache.flush</code></quote>
841          (<span class="type">unsigned</span>)
842          <code class="literal">-w</code>
843        </span></dt><dd><p>Flush the specified thread-specific cache (tcache).  The
844        same considerations apply to this interface as to <a class="link" href="#thread.tcache.flush"><quote><code class="mallctl">thread.tcache.flush</code></quote></a>,
845        except that the tcache will never be automatically discarded.
846        </p></dd><dt><a name="tcache.destroy"/><span class="term">
847          <quote><code class="mallctl">tcache.destroy</code></quote>
848          (<span class="type">unsigned</span>)
849          <code class="literal">-w</code>
850        </span></dt><dd><p>Flush the specified thread-specific cache (tcache) and
851        make the identifier available for use during a future tcache creation.
852        </p></dd><dt><a name="arena.i.initialized"/><span class="term">
853          <quote><code class="mallctl">arena.&lt;i&gt;.initialized</code></quote>
854          (<span class="type">bool</span>)
855          <code class="literal">r-</code>
856        </span></dt><dd><p>Get whether the specified arena's statistics are
857        initialized (i.e. the arena was initialized prior to the current epoch).
858        This interface can also be nominally used to query whether the merged
859        statistics corresponding to <code class="constant">MALLCTL_ARENAS_ALL</code> are
860        initialized (always true).</p></dd><dt><a name="arena.i.decay"/><span class="term">
861          <quote><code class="mallctl">arena.&lt;i&gt;.decay</code></quote>
862          (<span class="type">void</span>)
863          <code class="literal">--</code>
864        </span></dt><dd><p>Trigger decay-based purging of unused dirty/muzzy pages
865        for arena &lt;i&gt;, or for all arenas if &lt;i&gt; equals
866        <code class="constant">MALLCTL_ARENAS_ALL</code>.  The proportion of unused
867        dirty/muzzy pages to be purged depends on the current time; see <a class="link" href="#opt.dirty_decay_ms"><quote><code class="mallctl">opt.dirty_decay_ms</code></quote></a>
868        and <a class="link" href="#opt.muzzy_decay_ms"><quote><code class="mallctl">opt.muzy_decay_ms</code></quote></a>
869        for details.</p></dd><dt><a name="arena.i.purge"/><span class="term">
870          <quote><code class="mallctl">arena.&lt;i&gt;.purge</code></quote>
871          (<span class="type">void</span>)
872          <code class="literal">--</code>
873        </span></dt><dd><p>Purge all unused dirty pages for arena &lt;i&gt;, or for
874        all arenas if &lt;i&gt; equals <code class="constant">MALLCTL_ARENAS_ALL</code>.
875        </p></dd><dt><a name="arena.i.reset"/><span class="term">
876          <quote><code class="mallctl">arena.&lt;i&gt;.reset</code></quote>
877          (<span class="type">void</span>)
878          <code class="literal">--</code>
879        </span></dt><dd><p>Discard all of the arena's extant allocations.  This
880        interface can only be used with arenas explicitly created via <a class="link" href="#arenas.create"><quote><code class="mallctl">arenas.create</code></quote></a>.  None
881        of the arena's discarded/cached allocations may accessed afterward.  As
882        part of this requirement, all thread caches which were used to
883        allocate/deallocate in conjunction with the arena must be flushed
884        beforehand.</p></dd><dt><a name="arena.i.destroy"/><span class="term">
885          <quote><code class="mallctl">arena.&lt;i&gt;.destroy</code></quote>
886          (<span class="type">void</span>)
887          <code class="literal">--</code>
888        </span></dt><dd><p>Destroy the arena.  Discard all of the arena's extant
889        allocations using the same mechanism as for <a class="link" href="#arena.i.reset"><quote><code class="mallctl">arena.&lt;i&gt;.reset</code></quote></a>
890        (with all the same constraints and side effects), merge the arena stats
891        into those accessible at arena index
892        <code class="constant">MALLCTL_ARENAS_DESTROYED</code>, and then completely
893        discard all metadata associated with the arena.  Future calls to <a class="link" href="#arenas.create"><quote><code class="mallctl">arenas.create</code></quote></a> may
894        recycle the arena index.  Destruction will fail if any threads are
895        currently associated with the arena as a result of calls to <a class="link" href="#thread.arena"><quote><code class="mallctl">thread.arena</code></quote></a>.</p></dd><dt><a name="arena.i.dss"/><span class="term">
896          <quote><code class="mallctl">arena.&lt;i&gt;.dss</code></quote>
897          (<span class="type">const char *</span>)
898          <code class="literal">rw</code>
899        </span></dt><dd><p>Set the precedence of dss allocation as related to mmap
900        allocation for arena &lt;i&gt;, or for all arenas if &lt;i&gt; equals
901        <code class="constant">MALLCTL_ARENAS_ALL</code>.  See <a class="link" href="#opt.dss"><quote><code class="mallctl">opt.dss</code></quote></a> for supported
902        settings.</p></dd><dt><a name="arena.i.dirty_decay_ms"/><span class="term">
903          <quote><code class="mallctl">arena.&lt;i&gt;.dirty_decay_ms</code></quote>
904          (<span class="type">ssize_t</span>)
905          <code class="literal">rw</code>
906        </span></dt><dd><p>Current per-arena approximate time in milliseconds from
907        the creation of a set of unused dirty pages until an equivalent set of
908        unused dirty pages is purged and/or reused.  Each time this interface is
909        set, all currently unused dirty pages are considered to have fully
910        decayed, which causes immediate purging of all unused dirty pages unless
911        the decay time is set to -1 (i.e. purging disabled).  See <a class="link" href="#opt.dirty_decay_ms"><quote><code class="mallctl">opt.dirty_decay_ms</code></quote></a>
912        for additional information.</p></dd><dt><a name="arena.i.muzzy_decay_ms"/><span class="term">
913          <quote><code class="mallctl">arena.&lt;i&gt;.muzzy_decay_ms</code></quote>
914          (<span class="type">ssize_t</span>)
915          <code class="literal">rw</code>
916        </span></dt><dd><p>Current per-arena approximate time in milliseconds from
917        the creation of a set of unused muzzy pages until an equivalent set of
918        unused muzzy pages is purged and/or reused.  Each time this interface is
919        set, all currently unused muzzy pages are considered to have fully
920        decayed, which causes immediate purging of all unused muzzy pages unless
921        the decay time is set to -1 (i.e. purging disabled).  See <a class="link" href="#opt.muzzy_decay_ms"><quote><code class="mallctl">opt.muzzy_decay_ms</code></quote></a>
922        for additional information.</p></dd><dt><a name="arena.i.retain_grow_limit"/><span class="term">
923          <quote><code class="mallctl">arena.&lt;i&gt;.retain_grow_limit</code></quote>
924          (<span class="type">size_t</span>)
925          <code class="literal">rw</code>
926        </span></dt><dd><p>Maximum size to grow retained region (only relevant when
927        <a class="link" href="#opt.retain"><quote><code class="mallctl">opt.retain</code></quote></a> is
928        enabled).  This controls the maximum increment to expand virtual memory,
929        or allocation through <a class="link" href="#arena.i.extent_hooks"><quote><code class="mallctl">arena.&lt;i&gt;extent_hooks</code></quote></a>.
930        In particular, if customized extent hooks reserve physical memory
931        (e.g. 1G huge pages), this is useful to control the allocation hook's
932        input size.  The default is no limit.</p></dd><dt><a name="arena.i.extent_hooks"/><span class="term">
933          <quote><code class="mallctl">arena.&lt;i&gt;.extent_hooks</code></quote>
934          (<span class="type">extent_hooks_t *</span>)
935          <code class="literal">rw</code>
936        </span></dt><dd><p>Get or set the extent management hook functions for
937        arena &lt;i&gt;.  The functions must be capable of operating on all
938        extant extents associated with arena &lt;i&gt;, usually by passing
939        unknown extents to the replaced functions.  In practice, it is feasible
940        to control allocation for arenas explicitly created via <a class="link" href="#arenas.create"><quote><code class="mallctl">arenas.create</code></quote></a> such
941        that all extents originate from an application-supplied extent allocator
942        (by specifying the custom extent hook functions during arena creation),
943        but the automatically created arenas will have already created extents
944        prior to the application having an opportunity to take over extent
945        allocation.</p><pre class="programlisting">
946typedef extent_hooks_s extent_hooks_t;
947struct extent_hooks_s {
948	extent_alloc_t		*alloc;
949	extent_dalloc_t		*dalloc;
950	extent_destroy_t	*destroy;
951	extent_commit_t		*commit;
952	extent_decommit_t	*decommit;
953	extent_purge_t		*purge_lazy;
954	extent_purge_t		*purge_forced;
955	extent_split_t		*split;
956	extent_merge_t		*merge;
957};</pre><p>The <span class="type">extent_hooks_t</span> structure comprises function
958        pointers which are described individually below.  jemalloc uses these
959        functions to manage extent lifetime, which starts off with allocation of
960        mapped committed memory, in the simplest case followed by deallocation.
961        However, there are performance and platform reasons to retain extents
962        for later reuse.  Cleanup attempts cascade from deallocation to decommit
963        to forced purging to lazy purging, which gives the extent management
964        functions opportunities to reject the most permanent cleanup operations
965        in favor of less permanent (and often less costly) operations.  All
966        operations except allocation can be universally opted out of by setting
967        the hook pointers to <code class="constant">NULL</code>, or selectively opted out
968        of by returning failure.  Note that once the extent hook is set, the
969        structure is accessed directly by the associated arenas, so it must
970        remain valid for the entire lifetime of the arenas.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef void *<b class="fsfunc">(extent_alloc_t)</b>(</code></td><td>extent_hooks_t *<var class="pdparam">extent_hooks</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">new_addr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td> </td><td>bool *<var class="pdparam">zero</var>, </td></tr><tr><td> </td><td>bool *<var class="pdparam">commit</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>An extent allocation function conforms to the
971        <span class="type">extent_alloc_t</span> type and upon success returns a pointer to
972        <em class="parameter"><code>size</code></em> bytes of mapped memory on behalf of arena
973        <em class="parameter"><code>arena_ind</code></em> such that the extent's base address is
974        a multiple of <em class="parameter"><code>alignment</code></em>, as well as setting
975        <em class="parameter"><code>*zero</code></em> to indicate whether the extent is zeroed
976        and <em class="parameter"><code>*commit</code></em> to indicate whether the extent is
977        committed.  Upon error the function returns <code class="constant">NULL</code>
978        and leaves <em class="parameter"><code>*zero</code></em> and
979        <em class="parameter"><code>*commit</code></em> unmodified.  The
980        <em class="parameter"><code>size</code></em> parameter is always a multiple of the page
981        size.  The <em class="parameter"><code>alignment</code></em> parameter is always a power
982        of two at least as large as the page size.  Zeroing is mandatory if
983        <em class="parameter"><code>*zero</code></em> is true upon function entry.  Committing is
984        mandatory if <em class="parameter"><code>*commit</code></em> is true upon function entry.
985        If <em class="parameter"><code>new_addr</code></em> is not <code class="constant">NULL</code>, the
986        returned pointer must be <em class="parameter"><code>new_addr</code></em> on success or
987        <code class="constant">NULL</code> on error.  Committed memory may be committed
988        in absolute terms as on a system that does not overcommit, or in
989        implicit terms as on a system that overcommits and satisfies physical
990        memory needs on demand via soft page faults.  Note that replacing the
991        default extent allocation function makes the arena's <a class="link" href="#arena.i.dss"><quote><code class="mallctl">arena.&lt;i&gt;.dss</code></quote></a>
992        setting irrelevant.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(extent_dalloc_t)</b>(</code></td><td>extent_hooks_t *<var class="pdparam">extent_hooks</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">addr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>
993        An extent deallocation function conforms to the
994        <span class="type">extent_dalloc_t</span> type and deallocates an extent at given
995        <em class="parameter"><code>addr</code></em> and <em class="parameter"><code>size</code></em> with
996        <em class="parameter"><code>committed</code></em>/decommited memory as indicated, on
997        behalf of arena <em class="parameter"><code>arena_ind</code></em>, returning false upon
998        success.  If the function returns true, this indicates opt-out from
999        deallocation; the virtual memory mapping associated with the extent
1000        remains mapped, in the same commit state, and available for future use,
1001        in which case it will be automatically retained for later reuse.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef void <b class="fsfunc">(extent_destroy_t)</b>(</code></td><td>extent_hooks_t *<var class="pdparam">extent_hooks</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">addr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>
1002        An extent destruction function conforms to the
1003        <span class="type">extent_destroy_t</span> type and unconditionally destroys an
1004        extent at given <em class="parameter"><code>addr</code></em> and
1005        <em class="parameter"><code>size</code></em> with
1006        <em class="parameter"><code>committed</code></em>/decommited memory as indicated, on
1007        behalf of arena <em class="parameter"><code>arena_ind</code></em>.  This function may be
1008        called to destroy retained extents during arena destruction (see <a class="link" href="#arena.i.destroy"><quote><code class="mallctl">arena.&lt;i&gt;.destroy</code></quote></a>).</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(extent_commit_t)</b>(</code></td><td>extent_hooks_t *<var class="pdparam">extent_hooks</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">addr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>An extent commit function conforms to the
1009        <span class="type">extent_commit_t</span> type and commits zeroed physical memory to
1010        back pages within an extent at given <em class="parameter"><code>addr</code></em> and
1011        <em class="parameter"><code>size</code></em> at <em class="parameter"><code>offset</code></em> bytes,
1012        extending for <em class="parameter"><code>length</code></em> on behalf of arena
1013        <em class="parameter"><code>arena_ind</code></em>, returning false upon success.
1014        Committed memory may be committed in absolute terms as on a system that
1015        does not overcommit, or in implicit terms as on a system that
1016        overcommits and satisfies physical memory needs on demand via soft page
1017        faults. If the function returns true, this indicates insufficient
1018        physical memory to satisfy the request.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(extent_decommit_t)</b>(</code></td><td>extent_hooks_t *<var class="pdparam">extent_hooks</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">addr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>An extent decommit function conforms to the
1019        <span class="type">extent_decommit_t</span> type and decommits any physical memory
1020        that is backing pages within an extent at given
1021        <em class="parameter"><code>addr</code></em> and <em class="parameter"><code>size</code></em> at
1022        <em class="parameter"><code>offset</code></em> bytes, extending for
1023        <em class="parameter"><code>length</code></em> on behalf of arena
1024        <em class="parameter"><code>arena_ind</code></em>, returning false upon success, in which
1025        case the pages will be committed via the extent commit function before
1026        being reused.  If the function returns true, this indicates opt-out from
1027        decommit; the memory remains committed and available for future use, in
1028        which case it will be automatically retained for later reuse.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(extent_purge_t)</b>(</code></td><td>extent_hooks_t *<var class="pdparam">extent_hooks</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">addr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">offset</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">length</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>An extent purge function conforms to the
1029        <span class="type">extent_purge_t</span> type and discards physical pages
1030        within the virtual memory mapping associated with an extent at given
1031        <em class="parameter"><code>addr</code></em> and <em class="parameter"><code>size</code></em> at
1032        <em class="parameter"><code>offset</code></em> bytes, extending for
1033        <em class="parameter"><code>length</code></em> on behalf of arena
1034        <em class="parameter"><code>arena_ind</code></em>.  A lazy extent purge function (e.g.
1035        implemented via
1036        <code class="function">madvise(<em class="parameter"><code>...</code></em><em class="parameter"><code><code class="constant">MADV_FREE</code></code></em>)</code>)
1037        can delay purging indefinitely and leave the pages within the purged
1038        virtual memory range in an indeterminite state, whereas a forced extent
1039        purge function immediately purges, and the pages within the virtual
1040        memory range will be zero-filled the next time they are accessed.  If
1041        the function returns true, this indicates failure to purge.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(extent_split_t)</b>(</code></td><td>extent_hooks_t *<var class="pdparam">extent_hooks</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">addr</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_a</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_b</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>An extent split function conforms to the
1042        <span class="type">extent_split_t</span> type and optionally splits an extent at
1043        given <em class="parameter"><code>addr</code></em> and <em class="parameter"><code>size</code></em> into
1044        two adjacent extents, the first of <em class="parameter"><code>size_a</code></em> bytes,
1045        and the second of <em class="parameter"><code>size_b</code></em> bytes, operating on
1046        <em class="parameter"><code>committed</code></em>/decommitted memory as indicated, on
1047        behalf of arena <em class="parameter"><code>arena_ind</code></em>, returning false upon
1048        success.  If the function returns true, this indicates that the extent
1049        remains unsplit and therefore should continue to be operated on as a
1050        whole.</p><div class="funcsynopsis"><table border="0" class="funcprototype-table" summary="Function synopsis" style="cellspacing: 0; cellpadding: 0;"><tr><td><code class="funcdef">typedef bool <b class="fsfunc">(extent_merge_t)</b>(</code></td><td>extent_hooks_t *<var class="pdparam">extent_hooks</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">addr_a</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_a</var>, </td></tr><tr><td> </td><td>void *<var class="pdparam">addr_b</var>, </td></tr><tr><td> </td><td>size_t <var class="pdparam">size_b</var>, </td></tr><tr><td> </td><td>bool <var class="pdparam">committed</var>, </td></tr><tr><td> </td><td>unsigned <var class="pdparam">arena_ind</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer"> </div></div><div class="literallayout"><p/></div><p>An extent merge function conforms to the
1051        <span class="type">extent_merge_t</span> type and optionally merges adjacent extents,
1052        at given <em class="parameter"><code>addr_a</code></em> and <em class="parameter"><code>size_a</code></em>
1053        with given <em class="parameter"><code>addr_b</code></em> and
1054        <em class="parameter"><code>size_b</code></em> into one contiguous extent, operating on
1055        <em class="parameter"><code>committed</code></em>/decommitted memory as indicated, on
1056        behalf of arena <em class="parameter"><code>arena_ind</code></em>, returning false upon
1057        success.  If the function returns true, this indicates that the extents
1058        remain distinct mappings and therefore should continue to be operated on
1059        independently.</p></dd><dt><a name="arenas.narenas"/><span class="term">
1060          <quote><code class="mallctl">arenas.narenas</code></quote>
1061          (<span class="type">unsigned</span>)
1062          <code class="literal">r-</code>
1063        </span></dt><dd><p>Current limit on number of arenas.</p></dd><dt><a name="arenas.dirty_decay_ms"/><span class="term">
1064          <quote><code class="mallctl">arenas.dirty_decay_ms</code></quote>
1065          (<span class="type">ssize_t</span>)
1066          <code class="literal">rw</code>
1067        </span></dt><dd><p>Current default per-arena approximate time in
1068        milliseconds from the creation of a set of unused dirty pages until an
1069        equivalent set of unused dirty pages is purged and/or reused, used to
1070        initialize <a class="link" href="#arena.i.dirty_decay_ms"><quote><code class="mallctl">arena.&lt;i&gt;.dirty_decay_ms</code></quote></a>
1071        during arena creation.  See <a class="link" href="#opt.dirty_decay_ms"><quote><code class="mallctl">opt.dirty_decay_ms</code></quote></a>
1072        for additional information.</p></dd><dt><a name="arenas.muzzy_decay_ms"/><span class="term">
1073          <quote><code class="mallctl">arenas.muzzy_decay_ms</code></quote>
1074          (<span class="type">ssize_t</span>)
1075          <code class="literal">rw</code>
1076        </span></dt><dd><p>Current default per-arena approximate time in
1077        milliseconds from the creation of a set of unused muzzy pages until an
1078        equivalent set of unused muzzy pages is purged and/or reused, used to
1079        initialize <a class="link" href="#arena.i.muzzy_decay_ms"><quote><code class="mallctl">arena.&lt;i&gt;.muzzy_decay_ms</code></quote></a>
1080        during arena creation.  See <a class="link" href="#opt.muzzy_decay_ms"><quote><code class="mallctl">opt.muzzy_decay_ms</code></quote></a>
1081        for additional information.</p></dd><dt><a name="arenas.quantum"/><span class="term">
1082          <quote><code class="mallctl">arenas.quantum</code></quote>
1083          (<span class="type">size_t</span>)
1084          <code class="literal">r-</code>
1085        </span></dt><dd><p>Quantum size.</p></dd><dt><a name="arenas.page"/><span class="term">
1086          <quote><code class="mallctl">arenas.page</code></quote>
1087          (<span class="type">size_t</span>)
1088          <code class="literal">r-</code>
1089        </span></dt><dd><p>Page size.</p></dd><dt><a name="arenas.tcache_max"/><span class="term">
1090          <quote><code class="mallctl">arenas.tcache_max</code></quote>
1091          (<span class="type">size_t</span>)
1092          <code class="literal">r-</code>
1093        </span></dt><dd><p>Maximum thread-cached size class.</p></dd><dt><a name="arenas.nbins"/><span class="term">
1094          <quote><code class="mallctl">arenas.nbins</code></quote>
1095          (<span class="type">unsigned</span>)
1096          <code class="literal">r-</code>
1097        </span></dt><dd><p>Number of bin size classes.</p></dd><dt><a name="arenas.nhbins"/><span class="term">
1098          <quote><code class="mallctl">arenas.nhbins</code></quote>
1099          (<span class="type">unsigned</span>)
1100          <code class="literal">r-</code>
1101        </span></dt><dd><p>Total number of thread cache bin size
1102        classes.</p></dd><dt><a name="arenas.bin.i.size"/><span class="term">
1103          <quote><code class="mallctl">arenas.bin.&lt;i&gt;.size</code></quote>
1104          (<span class="type">size_t</span>)
1105          <code class="literal">r-</code>
1106        </span></dt><dd><p>Maximum size supported by size class.</p></dd><dt><a name="arenas.bin.i.nregs"/><span class="term">
1107          <quote><code class="mallctl">arenas.bin.&lt;i&gt;.nregs</code></quote>
1108          (<span class="type">uint32_t</span>)
1109          <code class="literal">r-</code>
1110        </span></dt><dd><p>Number of regions per slab.</p></dd><dt><a name="arenas.bin.i.slab_size"/><span class="term">
1111          <quote><code class="mallctl">arenas.bin.&lt;i&gt;.slab_size</code></quote>
1112          (<span class="type">size_t</span>)
1113          <code class="literal">r-</code>
1114        </span></dt><dd><p>Number of bytes per slab.</p></dd><dt><a name="arenas.nlextents"/><span class="term">
1115          <quote><code class="mallctl">arenas.nlextents</code></quote>
1116          (<span class="type">unsigned</span>)
1117          <code class="literal">r-</code>
1118        </span></dt><dd><p>Total number of large size classes.</p></dd><dt><a name="arenas.lextent.i.size"/><span class="term">
1119          <quote><code class="mallctl">arenas.lextent.&lt;i&gt;.size</code></quote>
1120          (<span class="type">size_t</span>)
1121          <code class="literal">r-</code>
1122        </span></dt><dd><p>Maximum size supported by this large size
1123        class.</p></dd><dt><a name="arenas.create"/><span class="term">
1124          <quote><code class="mallctl">arenas.create</code></quote>
1125          (<span class="type">unsigned</span>, <span class="type">extent_hooks_t *</span>)
1126          <code class="literal">rw</code>
1127        </span></dt><dd><p>Explicitly create a new arena outside the range of
1128        automatically managed arenas, with optionally specified extent hooks,
1129        and return the new arena index.</p></dd><dt><a name="arenas.lookup"/><span class="term">
1130          <quote><code class="mallctl">arenas.lookup</code></quote>
1131          (<span class="type">unsigned</span>, <span class="type">void*</span>)
1132          <code class="literal">rw</code>
1133        </span></dt><dd><p>Index of the arena to which an allocation belongs to.</p></dd><dt><a name="prof.thread_active_init"/><span class="term">
1134          <quote><code class="mallctl">prof.thread_active_init</code></quote>
1135          (<span class="type">bool</span>)
1136          <code class="literal">rw</code>
1137          [<code class="option">--enable-prof</code>]
1138        </span></dt><dd><p>Control the initial setting for <a class="link" href="#thread.prof.active"><quote><code class="mallctl">thread.prof.active</code></quote></a>
1139        in newly created threads.  See the <a class="link" href="#opt.prof_thread_active_init"><quote><code class="mallctl">opt.prof_thread_active_init</code></quote></a>
1140        option for additional information.</p></dd><dt><a name="prof.active"/><span class="term">
1141          <quote><code class="mallctl">prof.active</code></quote>
1142          (<span class="type">bool</span>)
1143          <code class="literal">rw</code>
1144          [<code class="option">--enable-prof</code>]
1145        </span></dt><dd><p>Control whether sampling is currently active.  See the
1146        <a class="link" href="#opt.prof_active"><quote><code class="mallctl">opt.prof_active</code></quote></a>
1147        option for additional information, as well as the interrelated <a class="link" href="#thread.prof.active"><quote><code class="mallctl">thread.prof.active</code></quote></a>
1148        mallctl.</p></dd><dt><a name="prof.dump"/><span class="term">
1149          <quote><code class="mallctl">prof.dump</code></quote>
1150          (<span class="type">const char *</span>)
1151          <code class="literal">-w</code>
1152          [<code class="option">--enable-prof</code>]
1153        </span></dt><dd><p>Dump a memory profile to the specified file, or if NULL
1154        is specified, to a file according to the pattern
1155        <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.m&lt;mseq&gt;.heap</code>,
1156        where <code class="literal">&lt;prefix&gt;</code> is controlled by the
1157        <a class="link" href="#opt.prof_prefix"><quote><code class="mallctl">opt.prof_prefix</code></quote></a>
1158        option.</p></dd><dt><a name="prof.gdump"/><span class="term">
1159          <quote><code class="mallctl">prof.gdump</code></quote>
1160          (<span class="type">bool</span>)
1161          <code class="literal">rw</code>
1162          [<code class="option">--enable-prof</code>]
1163        </span></dt><dd><p>When enabled, trigger a memory profile dump every time
1164        the total virtual memory exceeds the previous maximum.  Profiles are
1165        dumped to files named according to the pattern
1166        <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.u&lt;useq&gt;.heap</code>,
1167        where <code class="literal">&lt;prefix&gt;</code> is controlled by the <a class="link" href="#opt.prof_prefix"><quote><code class="mallctl">opt.prof_prefix</code></quote></a>
1168        option.</p></dd><dt><a name="prof.reset"/><span class="term">
1169          <quote><code class="mallctl">prof.reset</code></quote>
1170          (<span class="type">size_t</span>)
1171          <code class="literal">-w</code>
1172          [<code class="option">--enable-prof</code>]
1173        </span></dt><dd><p>Reset all memory profile statistics, and optionally
1174        update the sample rate (see <a class="link" href="#opt.lg_prof_sample"><quote><code class="mallctl">opt.lg_prof_sample</code></quote></a>
1175        and <a class="link" href="#prof.lg_sample"><quote><code class="mallctl">prof.lg_sample</code></quote></a>).
1176        </p></dd><dt><a name="prof.lg_sample"/><span class="term">
1177          <quote><code class="mallctl">prof.lg_sample</code></quote>
1178          (<span class="type">size_t</span>)
1179          <code class="literal">r-</code>
1180          [<code class="option">--enable-prof</code>]
1181        </span></dt><dd><p>Get the current sample rate (see <a class="link" href="#opt.lg_prof_sample"><quote><code class="mallctl">opt.lg_prof_sample</code></quote></a>).
1182        </p></dd><dt><a name="prof.interval"/><span class="term">
1183          <quote><code class="mallctl">prof.interval</code></quote>
1184          (<span class="type">uint64_t</span>)
1185          <code class="literal">r-</code>
1186          [<code class="option">--enable-prof</code>]
1187        </span></dt><dd><p>Average number of bytes allocated between
1188        interval-based profile dumps.  See the
1189        <a class="link" href="#opt.lg_prof_interval"><quote><code class="mallctl">opt.lg_prof_interval</code></quote></a>
1190        option for additional information.</p></dd><dt><a name="stats.allocated"/><span class="term">
1191          <quote><code class="mallctl">stats.allocated</code></quote>
1192          (<span class="type">size_t</span>)
1193          <code class="literal">r-</code>
1194          [<code class="option">--enable-stats</code>]
1195        </span></dt><dd><p>Total number of bytes allocated by the
1196        application.</p></dd><dt><a name="stats.active"/><span class="term">
1197          <quote><code class="mallctl">stats.active</code></quote>
1198          (<span class="type">size_t</span>)
1199          <code class="literal">r-</code>
1200          [<code class="option">--enable-stats</code>]
1201        </span></dt><dd><p>Total number of bytes in active pages allocated by the
1202        application.  This is a multiple of the page size, and greater than or
1203        equal to <a class="link" href="#stats.allocated"><quote><code class="mallctl">stats.allocated</code></quote></a>.
1204        This does not include <a class="link" href="#stats.arenas.i.pdirty">
1205        <quote><code class="mallctl">stats.arenas.&lt;i&gt;.pdirty</code></quote></a>,
1206        <a class="link" href="#stats.arenas.i.pmuzzy">
1207        <quote><code class="mallctl">stats.arenas.&lt;i&gt;.pmuzzy</code></quote></a>, nor pages
1208        entirely devoted to allocator metadata.</p></dd><dt><a name="stats.metadata"/><span class="term">
1209          <quote><code class="mallctl">stats.metadata</code></quote>
1210          (<span class="type">size_t</span>)
1211          <code class="literal">r-</code>
1212          [<code class="option">--enable-stats</code>]
1213        </span></dt><dd><p>Total number of bytes dedicated to metadata, which
1214        comprise base allocations used for bootstrap-sensitive allocator
1215        metadata structures (see <a class="link" href="#stats.arenas.i.base"><quote><code class="mallctl">stats.arenas.&lt;i&gt;.base</code></quote></a>)
1216        and internal allocations (see <a class="link" href="#stats.arenas.i.internal"><quote><code class="mallctl">stats.arenas.&lt;i&gt;.internal</code></quote></a>).
1217        Transparent huge page (enabled with <a class="link" href="#opt.metadata_thp">opt.metadata_thp</a>) usage is not
1218        considered.</p></dd><dt><a name="stats.metadata_thp"/><span class="term">
1219          <quote><code class="mallctl">stats.metadata_thp</code></quote>
1220          (<span class="type">size_t</span>)
1221          <code class="literal">r-</code>
1222          [<code class="option">--enable-stats</code>]
1223        </span></dt><dd><p>Number of transparent huge pages (THP) used for
1224        metadata.  See <a class="link" href="#stats.metadata"><quote><code class="mallctl">stats.metadata</code></quote></a> and
1225        <a class="link" href="#opt.metadata_thp">opt.metadata_thp</a>) for
1226        details.</p></dd><dt><a name="stats.resident"/><span class="term">
1227          <quote><code class="mallctl">stats.resident</code></quote>
1228          (<span class="type">size_t</span>)
1229          <code class="literal">r-</code>
1230          [<code class="option">--enable-stats</code>]
1231        </span></dt><dd><p>Maximum number of bytes in physically resident data
1232        pages mapped by the allocator, comprising all pages dedicated to
1233        allocator metadata, pages backing active allocations, and unused dirty
1234        pages.  This is a maximum rather than precise because pages may not
1235        actually be physically resident if they correspond to demand-zeroed
1236        virtual memory that has not yet been touched.  This is a multiple of the
1237        page size, and is larger than <a class="link" href="#stats.active"><quote><code class="mallctl">stats.active</code></quote></a>.</p></dd><dt><a name="stats.mapped"/><span class="term">
1238          <quote><code class="mallctl">stats.mapped</code></quote>
1239          (<span class="type">size_t</span>)
1240          <code class="literal">r-</code>
1241          [<code class="option">--enable-stats</code>]
1242        </span></dt><dd><p>Total number of bytes in active extents mapped by the
1243        allocator.  This is larger than <a class="link" href="#stats.active"><quote><code class="mallctl">stats.active</code></quote></a>.  This
1244        does not include inactive extents, even those that contain unused dirty
1245        pages, which means that there is no strict ordering between this and
1246        <a class="link" href="#stats.resident"><quote><code class="mallctl">stats.resident</code></quote></a>.</p></dd><dt><a name="stats.retained"/><span class="term">
1247          <quote><code class="mallctl">stats.retained</code></quote>
1248          (<span class="type">size_t</span>)
1249          <code class="literal">r-</code>
1250          [<code class="option">--enable-stats</code>]
1251        </span></dt><dd><p>Total number of bytes in virtual memory mappings that
1252        were retained rather than being returned to the operating system via
1253        e.g. <span class="citerefentry"><span class="refentrytitle">munmap</span>(2)</span> or similar.  Retained virtual
1254        memory is typically untouched, decommitted, or purged, so it has no
1255        strongly associated physical memory (see <a class="link" href="#arena.i.extent_hooks">extent hooks</a> for details).
1256        Retained memory is excluded from mapped memory statistics, e.g. <a class="link" href="#stats.mapped"><quote><code class="mallctl">stats.mapped</code></quote></a>.
1257        </p></dd><dt><a name="stats.background_thread.num_threads"/><span class="term">
1258          <quote><code class="mallctl">stats.background_thread.num_threads</code></quote>
1259          (<span class="type">size_t</span>)
1260          <code class="literal">r-</code>
1261          [<code class="option">--enable-stats</code>]
1262        </span></dt><dd><p> Number of <a class="link" href="#background_thread">background
1263        threads</a> running currently.</p></dd><dt><a name="stats.background_thread.num_runs"/><span class="term">
1264          <quote><code class="mallctl">stats.background_thread.num_runs</code></quote>
1265          (<span class="type">uint64_t</span>)
1266          <code class="literal">r-</code>
1267          [<code class="option">--enable-stats</code>]
1268        </span></dt><dd><p> Total number of runs from all <a class="link" href="#background_thread">background threads</a>.</p></dd><dt><a name="stats.background_thread.run_interval"/><span class="term">
1269          <quote><code class="mallctl">stats.background_thread.run_interval</code></quote>
1270          (<span class="type">uint64_t</span>)
1271          <code class="literal">r-</code>
1272          [<code class="option">--enable-stats</code>]
1273        </span></dt><dd><p> Average run interval in nanoseconds of <a class="link" href="#background_thread">background threads</a>.</p></dd><dt><a name="stats.mutexes.ctl"/><span class="term">
1274          <quote><code class="mallctl">stats.mutexes.ctl.{counter};</code></quote>
1275          (<span class="type">counter specific type</span>)
1276          <code class="literal">r-</code>
1277          [<code class="option">--enable-stats</code>]
1278        </span></dt><dd><p>Statistics on <code class="varname">ctl</code> mutex (global
1279        scope; mallctl related).  <quote><code class="mallctl">{counter}</code></quote> is one of the
1280        counters below:</p><dt><a name="mutex_counters"/></dt><dd><p><code class="varname">num_ops</code> (<span class="type">uint64_t</span>):
1281          Total number of lock acquisition operations on this mutex.</p><p><code class="varname">num_spin_acq</code> (<span class="type">uint64_t</span>): Number
1282	  of times the mutex was spin-acquired.  When the mutex is currently
1283	  locked and cannot be acquired immediately, a short period of
1284	  spin-retry within jemalloc will be performed.  Acquired through spin
1285	  generally means the contention was lightweight and not causing context
1286	  switches.</p><p><code class="varname">num_wait</code> (<span class="type">uint64_t</span>): Number of
1287	  times the mutex was wait-acquired, which means the mutex contention
1288	  was not solved by spin-retry, and blocking operation was likely
1289	  involved in order to acquire the mutex.  This event generally implies
1290	  higher cost / longer delay, and should be investigated if it happens
1291	  often.</p><p><code class="varname">max_wait_time</code> (<span class="type">uint64_t</span>):
1292	  Maximum length of time in nanoseconds spent on a single wait-acquired
1293	  lock operation.  Note that to avoid profiling overhead on the common
1294	  path, this does not consider spin-acquired cases.</p><p><code class="varname">total_wait_time</code> (<span class="type">uint64_t</span>):
1295	  Cumulative time in nanoseconds spent on wait-acquired lock operations.
1296	  Similarly, spin-acquired cases are not considered.</p><p><code class="varname">max_num_thds</code> (<span class="type">uint32_t</span>): Maximum
1297	  number of threads waiting on this mutex simultaneously.  Similarly,
1298	  spin-acquired cases are not considered.</p><p><code class="varname">num_owner_switch</code> (<span class="type">uint64_t</span>):
1299	  Number of times the current mutex owner is different from the previous
1300	  one.  This event does not generally imply an issue; rather it is an
1301	  indicator of how often the protected data are accessed by different
1302	  threads.
1303	  </p></dd></dd><dt><a name="stats.mutexes.background_thread"/><span class="term">
1304          <quote><code class="mallctl">stats.mutexes.background_thread.{counter}</code></quote>
1305	  (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1306          [<code class="option">--enable-stats</code>]
1307        </span></dt><dd><p>Statistics on <code class="varname">background_thread</code> mutex
1308        (global scope; <a class="link" href="#background_thread"><quote><code class="mallctl">background_thread</code></quote></a>
1309        related).  <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1310        counters</a>.</p></dd><dt><a name="stats.mutexes.prof"/><span class="term">
1311          <quote><code class="mallctl">stats.mutexes.prof.{counter}</code></quote>
1312	  (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1313          [<code class="option">--enable-stats</code>]
1314        </span></dt><dd><p>Statistics on <code class="varname">prof</code> mutex (global
1315        scope; profiling related).  <quote><code class="mallctl">{counter}</code></quote> is one of the
1316        counters in <a class="link" href="#mutex_counters">mutex profiling
1317        counters</a>.</p></dd><dt><a name="stats.mutexes.reset"/><span class="term">
1318          <quote><code class="mallctl">stats.mutexes.reset</code></quote>
1319	  (<span class="type">void</span>) <code class="literal">--</code>
1320          [<code class="option">--enable-stats</code>]
1321        </span></dt><dd><p>Reset all mutex profile statistics, including global
1322        mutexes, arena mutexes and bin mutexes.</p></dd><dt><a name="stats.arenas.i.dss"/><span class="term">
1323          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.dss</code></quote>
1324          (<span class="type">const char *</span>)
1325          <code class="literal">r-</code>
1326        </span></dt><dd><p>dss (<span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>) allocation precedence as
1327        related to <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> allocation.  See <a class="link" href="#opt.dss"><quote><code class="mallctl">opt.dss</code></quote></a> for details.
1328        </p></dd><dt><a name="stats.arenas.i.dirty_decay_ms"/><span class="term">
1329          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.dirty_decay_ms</code></quote>
1330          (<span class="type">ssize_t</span>)
1331          <code class="literal">r-</code>
1332        </span></dt><dd><p>Approximate time in milliseconds from the creation of a
1333        set of unused dirty pages until an equivalent set of unused dirty pages
1334        is purged and/or reused.  See <a class="link" href="#opt.dirty_decay_ms"><quote><code class="mallctl">opt.dirty_decay_ms</code></quote></a>
1335        for details.</p></dd><dt><a name="stats.arenas.i.muzzy_decay_ms"/><span class="term">
1336          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.muzzy_decay_ms</code></quote>
1337          (<span class="type">ssize_t</span>)
1338          <code class="literal">r-</code>
1339        </span></dt><dd><p>Approximate time in milliseconds from the creation of a
1340        set of unused muzzy pages until an equivalent set of unused muzzy pages
1341        is purged and/or reused.  See <a class="link" href="#opt.muzzy_decay_ms"><quote><code class="mallctl">opt.muzzy_decay_ms</code></quote></a>
1342        for details.</p></dd><dt><a name="stats.arenas.i.nthreads"/><span class="term">
1343          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.nthreads</code></quote>
1344          (<span class="type">unsigned</span>)
1345          <code class="literal">r-</code>
1346        </span></dt><dd><p>Number of threads currently assigned to
1347        arena.</p></dd><dt><a name="stats.arenas.i.uptime"/><span class="term">
1348          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.uptime</code></quote>
1349          (<span class="type">uint64_t</span>)
1350          <code class="literal">r-</code>
1351        </span></dt><dd><p>Time elapsed (in nanoseconds) since the arena was
1352        created.  If &lt;i&gt; equals <code class="constant">0</code> or
1353        <code class="constant">MALLCTL_ARENAS_ALL</code>, this is the uptime since malloc
1354        initialization.</p></dd><dt><a name="stats.arenas.i.pactive"/><span class="term">
1355          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.pactive</code></quote>
1356          (<span class="type">size_t</span>)
1357          <code class="literal">r-</code>
1358        </span></dt><dd><p>Number of pages in active extents.</p></dd><dt><a name="stats.arenas.i.pdirty"/><span class="term">
1359          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.pdirty</code></quote>
1360          (<span class="type">size_t</span>)
1361          <code class="literal">r-</code>
1362        </span></dt><dd><p>Number of pages within unused extents that are
1363        potentially dirty, and for which <code class="function">madvise()</code> or
1364        similar has not been called.  See <a class="link" href="#opt.dirty_decay_ms"><quote><code class="mallctl">opt.dirty_decay_ms</code></quote></a>
1365        for a description of dirty pages.</p></dd><dt><a name="stats.arenas.i.pmuzzy"/><span class="term">
1366          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.pmuzzy</code></quote>
1367          (<span class="type">size_t</span>)
1368          <code class="literal">r-</code>
1369        </span></dt><dd><p>Number of pages within unused extents that are muzzy.
1370        See <a class="link" href="#opt.muzzy_decay_ms"><quote><code class="mallctl">opt.muzzy_decay_ms</code></quote></a>
1371        for a description of muzzy pages.</p></dd><dt><a name="stats.arenas.i.mapped"/><span class="term">
1372          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mapped</code></quote>
1373          (<span class="type">size_t</span>)
1374          <code class="literal">r-</code>
1375          [<code class="option">--enable-stats</code>]
1376        </span></dt><dd><p>Number of mapped bytes.</p></dd><dt><a name="stats.arenas.i.retained"/><span class="term">
1377          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.retained</code></quote>
1378          (<span class="type">size_t</span>)
1379          <code class="literal">r-</code>
1380          [<code class="option">--enable-stats</code>]
1381        </span></dt><dd><p>Number of retained bytes.  See <a class="link" href="#stats.retained"><quote><code class="mallctl">stats.retained</code></quote></a> for
1382        details.</p></dd><dt><a name="stats.arenas.i.base"/><span class="term">
1383          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.base</code></quote>
1384          (<span class="type">size_t</span>)
1385          <code class="literal">r-</code>
1386          [<code class="option">--enable-stats</code>]
1387        </span></dt><dd><p>
1388        Number of bytes dedicated to bootstrap-sensitive allocator metadata
1389        structures.</p></dd><dt><a name="stats.arenas.i.internal"/><span class="term">
1390          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.internal</code></quote>
1391          (<span class="type">size_t</span>)
1392          <code class="literal">r-</code>
1393          [<code class="option">--enable-stats</code>]
1394        </span></dt><dd><p>Number of bytes dedicated to internal allocations.
1395        Internal allocations differ from application-originated allocations in
1396        that they are for internal use, and that they are omitted from heap
1397        profiles.</p></dd><dt><a name="stats.arenas.i.metadata_thp"/><span class="term">
1398          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.metadata_thp</code></quote>
1399          (<span class="type">size_t</span>)
1400          <code class="literal">r-</code>
1401          [<code class="option">--enable-stats</code>]
1402        </span></dt><dd><p>Number of transparent huge pages (THP) used for
1403        metadata.  See <a class="link" href="#opt.metadata_thp">opt.metadata_thp</a>
1404        for details.</p></dd><dt><a name="stats.arenas.i.resident"/><span class="term">
1405          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.resident</code></quote>
1406          (<span class="type">size_t</span>)
1407          <code class="literal">r-</code>
1408          [<code class="option">--enable-stats</code>]
1409        </span></dt><dd><p>Maximum number of bytes in physically resident data
1410        pages mapped by the arena, comprising all pages dedicated to allocator
1411        metadata, pages backing active allocations, and unused dirty pages.
1412        This is a maximum rather than precise because pages may not actually be
1413        physically resident if they correspond to demand-zeroed virtual memory
1414        that has not yet been touched.  This is a multiple of the page
1415        size.</p></dd><dt><a name="stats.arenas.i.dirty_npurge"/><span class="term">
1416          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.dirty_npurge</code></quote>
1417          (<span class="type">uint64_t</span>)
1418          <code class="literal">r-</code>
1419          [<code class="option">--enable-stats</code>]
1420        </span></dt><dd><p>Number of dirty page purge sweeps performed.
1421        </p></dd><dt><a name="stats.arenas.i.dirty_nmadvise"/><span class="term">
1422          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.dirty_nmadvise</code></quote>
1423          (<span class="type">uint64_t</span>)
1424          <code class="literal">r-</code>
1425          [<code class="option">--enable-stats</code>]
1426        </span></dt><dd><p>Number of <code class="function">madvise()</code> or similar
1427        calls made to purge dirty pages.</p></dd><dt><a name="stats.arenas.i.dirty_purged"/><span class="term">
1428          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.dirty_purged</code></quote>
1429          (<span class="type">uint64_t</span>)
1430          <code class="literal">r-</code>
1431          [<code class="option">--enable-stats</code>]
1432        </span></dt><dd><p>Number of dirty pages purged.</p></dd><dt><a name="stats.arenas.i.muzzy_npurge"/><span class="term">
1433          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.muzzy_npurge</code></quote>
1434          (<span class="type">uint64_t</span>)
1435          <code class="literal">r-</code>
1436          [<code class="option">--enable-stats</code>]
1437        </span></dt><dd><p>Number of muzzy page purge sweeps performed.
1438        </p></dd><dt><a name="stats.arenas.i.muzzy_nmadvise"/><span class="term">
1439          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.muzzy_nmadvise</code></quote>
1440          (<span class="type">uint64_t</span>)
1441          <code class="literal">r-</code>
1442          [<code class="option">--enable-stats</code>]
1443        </span></dt><dd><p>Number of <code class="function">madvise()</code> or similar
1444        calls made to purge muzzy pages.</p></dd><dt><a name="stats.arenas.i.muzzy_purged"/><span class="term">
1445          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.muzzy_purged</code></quote>
1446          (<span class="type">uint64_t</span>)
1447          <code class="literal">r-</code>
1448          [<code class="option">--enable-stats</code>]
1449        </span></dt><dd><p>Number of muzzy pages purged.</p></dd><dt><a name="stats.arenas.i.small.allocated"/><span class="term">
1450          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.small.allocated</code></quote>
1451          (<span class="type">size_t</span>)
1452          <code class="literal">r-</code>
1453          [<code class="option">--enable-stats</code>]
1454        </span></dt><dd><p>Number of bytes currently allocated by small objects.
1455        </p></dd><dt><a name="stats.arenas.i.small.nmalloc"/><span class="term">
1456          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.small.nmalloc</code></quote>
1457          (<span class="type">uint64_t</span>)
1458          <code class="literal">r-</code>
1459          [<code class="option">--enable-stats</code>]
1460        </span></dt><dd><p>Cumulative number of times a small allocation was
1461        requested from the arena's bins, whether to fill the relevant tcache if
1462        <a class="link" href="#opt.tcache"><quote><code class="mallctl">opt.tcache</code></quote></a> is
1463        enabled, or to directly satisfy an allocation request
1464        otherwise.</p></dd><dt><a name="stats.arenas.i.small.ndalloc"/><span class="term">
1465          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.small.ndalloc</code></quote>
1466          (<span class="type">uint64_t</span>)
1467          <code class="literal">r-</code>
1468          [<code class="option">--enable-stats</code>]
1469        </span></dt><dd><p>Cumulative number of times a small allocation was
1470        returned to the arena's bins, whether to flush the relevant tcache if
1471        <a class="link" href="#opt.tcache"><quote><code class="mallctl">opt.tcache</code></quote></a> is
1472        enabled, or to directly deallocate an allocation
1473        otherwise.</p></dd><dt><a name="stats.arenas.i.small.nrequests"/><span class="term">
1474          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.small.nrequests</code></quote>
1475          (<span class="type">uint64_t</span>)
1476          <code class="literal">r-</code>
1477          [<code class="option">--enable-stats</code>]
1478        </span></dt><dd><p>Cumulative number of allocation requests satisfied by
1479        all bin size classes.</p></dd><dt><a name="stats.arenas.i.large.allocated"/><span class="term">
1480          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.large.allocated</code></quote>
1481          (<span class="type">size_t</span>)
1482          <code class="literal">r-</code>
1483          [<code class="option">--enable-stats</code>]
1484        </span></dt><dd><p>Number of bytes currently allocated by large objects.
1485        </p></dd><dt><a name="stats.arenas.i.large.nmalloc"/><span class="term">
1486          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.large.nmalloc</code></quote>
1487          (<span class="type">uint64_t</span>)
1488          <code class="literal">r-</code>
1489          [<code class="option">--enable-stats</code>]
1490        </span></dt><dd><p>Cumulative number of times a large extent was allocated
1491        from the arena, whether to fill the relevant tcache if <a class="link" href="#opt.tcache"><quote><code class="mallctl">opt.tcache</code></quote></a> is enabled and
1492        the size class is within the range being cached, or to directly satisfy
1493        an allocation request otherwise.</p></dd><dt><a name="stats.arenas.i.large.ndalloc"/><span class="term">
1494          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.large.ndalloc</code></quote>
1495          (<span class="type">uint64_t</span>)
1496          <code class="literal">r-</code>
1497          [<code class="option">--enable-stats</code>]
1498        </span></dt><dd><p>Cumulative number of times a large extent was returned
1499        to the arena, whether to flush the relevant tcache if <a class="link" href="#opt.tcache"><quote><code class="mallctl">opt.tcache</code></quote></a> is enabled and
1500        the size class is within the range being cached, or to directly
1501        deallocate an allocation otherwise.</p></dd><dt><a name="stats.arenas.i.large.nrequests"/><span class="term">
1502          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.large.nrequests</code></quote>
1503          (<span class="type">uint64_t</span>)
1504          <code class="literal">r-</code>
1505          [<code class="option">--enable-stats</code>]
1506        </span></dt><dd><p>Cumulative number of allocation requests satisfied by
1507        all large size classes.</p></dd><dt><a name="stats.arenas.i.bins.j.nmalloc"/><span class="term">
1508          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nmalloc</code></quote>
1509          (<span class="type">uint64_t</span>)
1510          <code class="literal">r-</code>
1511          [<code class="option">--enable-stats</code>]
1512        </span></dt><dd><p>Cumulative number of times a bin region of the
1513        corresponding size class was allocated from the arena, whether to fill
1514        the relevant tcache if <a class="link" href="#opt.tcache"><quote><code class="mallctl">opt.tcache</code></quote></a> is enabled, or
1515        to directly satisfy an allocation request otherwise.</p></dd><dt><a name="stats.arenas.i.bins.j.ndalloc"/><span class="term">
1516          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.ndalloc</code></quote>
1517          (<span class="type">uint64_t</span>)
1518          <code class="literal">r-</code>
1519          [<code class="option">--enable-stats</code>]
1520        </span></dt><dd><p>Cumulative number of times a bin region of the
1521        corresponding size class was returned to the arena, whether to flush the
1522        relevant tcache if <a class="link" href="#opt.tcache"><quote><code class="mallctl">opt.tcache</code></quote></a> is enabled, or
1523        to directly deallocate an allocation otherwise.</p></dd><dt><a name="stats.arenas.i.bins.j.nrequests"/><span class="term">
1524          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nrequests</code></quote>
1525          (<span class="type">uint64_t</span>)
1526          <code class="literal">r-</code>
1527          [<code class="option">--enable-stats</code>]
1528        </span></dt><dd><p>Cumulative number of allocation requests satisfied by
1529        bin regions of the corresponding size class.</p></dd><dt><a name="stats.arenas.i.bins.j.curregs"/><span class="term">
1530          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.curregs</code></quote>
1531          (<span class="type">size_t</span>)
1532          <code class="literal">r-</code>
1533          [<code class="option">--enable-stats</code>]
1534        </span></dt><dd><p>Current number of regions for this size
1535        class.</p></dd><dt><a name="stats.arenas.i.bins.j.nfills"/><span class="term">
1536          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nfills</code></quote>
1537          (<span class="type">uint64_t</span>)
1538          <code class="literal">r-</code>
1539        </span></dt><dd><p>Cumulative number of tcache fills.</p></dd><dt><a name="stats.arenas.i.bins.j.nflushes"/><span class="term">
1540          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nflushes</code></quote>
1541          (<span class="type">uint64_t</span>)
1542          <code class="literal">r-</code>
1543        </span></dt><dd><p>Cumulative number of tcache flushes.</p></dd><dt><a name="stats.arenas.i.bins.j.nslabs"/><span class="term">
1544          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nslabs</code></quote>
1545          (<span class="type">uint64_t</span>)
1546          <code class="literal">r-</code>
1547          [<code class="option">--enable-stats</code>]
1548        </span></dt><dd><p>Cumulative number of slabs created.</p></dd><dt><a name="stats.arenas.i.bins.j.nreslabs"/><span class="term">
1549          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nreslabs</code></quote>
1550          (<span class="type">uint64_t</span>)
1551          <code class="literal">r-</code>
1552          [<code class="option">--enable-stats</code>]
1553        </span></dt><dd><p>Cumulative number of times the current slab from which
1554        to allocate changed.</p></dd><dt><a name="stats.arenas.i.bins.j.curslabs"/><span class="term">
1555          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.curslabs</code></quote>
1556          (<span class="type">size_t</span>)
1557          <code class="literal">r-</code>
1558          [<code class="option">--enable-stats</code>]
1559        </span></dt><dd><p>Current number of slabs.</p></dd><dt><a name="stats.arenas.i.bins.mutex"/><span class="term">
1560          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.mutex.{counter}</code></quote>
1561          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1562          [<code class="option">--enable-stats</code>]
1563        </span></dt><dd><p>Statistics on
1564        <code class="varname">arena.&lt;i&gt;.bins.&lt;j&gt;</code> mutex (arena bin
1565        scope; bin operation related).  <quote><code class="mallctl">{counter}</code></quote> is one of
1566        the counters in <a class="link" href="#mutex_counters">mutex profiling
1567        counters</a>.</p></dd><dt><a name="stats.arenas.i.lextents.j.nmalloc"/><span class="term">
1568          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lextents.&lt;j&gt;.nmalloc</code></quote>
1569          (<span class="type">uint64_t</span>)
1570          <code class="literal">r-</code>
1571          [<code class="option">--enable-stats</code>]
1572        </span></dt><dd><p>Cumulative number of times a large extent of the
1573        corresponding size class was allocated from the arena, whether to fill
1574        the relevant tcache if <a class="link" href="#opt.tcache"><quote><code class="mallctl">opt.tcache</code></quote></a> is enabled and
1575        the size class is within the range being cached, or to directly satisfy
1576        an allocation request otherwise.</p></dd><dt><a name="stats.arenas.i.lextents.j.ndalloc"/><span class="term">
1577          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lextents.&lt;j&gt;.ndalloc</code></quote>
1578          (<span class="type">uint64_t</span>)
1579          <code class="literal">r-</code>
1580          [<code class="option">--enable-stats</code>]
1581        </span></dt><dd><p>Cumulative number of times a large extent of the
1582        corresponding size class was returned to the arena, whether to flush the
1583        relevant tcache if <a class="link" href="#opt.tcache"><quote><code class="mallctl">opt.tcache</code></quote></a> is enabled and
1584        the size class is within the range being cached, or to directly
1585        deallocate an allocation otherwise.</p></dd><dt><a name="stats.arenas.i.lextents.j.nrequests"/><span class="term">
1586          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lextents.&lt;j&gt;.nrequests</code></quote>
1587          (<span class="type">uint64_t</span>)
1588          <code class="literal">r-</code>
1589          [<code class="option">--enable-stats</code>]
1590        </span></dt><dd><p>Cumulative number of allocation requests satisfied by
1591        large extents of the corresponding size class.</p></dd><dt><a name="stats.arenas.i.lextents.j.curlextents"/><span class="term">
1592          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.lextents.&lt;j&gt;.curlextents</code></quote>
1593          (<span class="type">size_t</span>)
1594          <code class="literal">r-</code>
1595          [<code class="option">--enable-stats</code>]
1596        </span></dt><dd><p>Current number of large allocations for this size class.
1597        </p></dd><dt><a name="stats.arenas.i.mutexes.large"/><span class="term">
1598          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.large.{counter}</code></quote>
1599          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1600          [<code class="option">--enable-stats</code>]
1601        </span></dt><dd><p>Statistics on <code class="varname">arena.&lt;i&gt;.large</code>
1602        mutex (arena scope; large allocation related).
1603        <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1604        counters</a>.</p></dd><dt><a name="stats.arenas.i.mutexes.extent_avail"/><span class="term">
1605          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.extent_avail.{counter}</code></quote>
1606          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1607          [<code class="option">--enable-stats</code>]
1608        </span></dt><dd><p>Statistics on <code class="varname">arena.&lt;i&gt;.extent_avail
1609        </code> mutex (arena scope; extent avail related).
1610        <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1611        counters</a>.</p></dd><dt><a name="stats.arenas.i.mutexes.extents_dirty"/><span class="term">
1612          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.extents_dirty.{counter}</code></quote>
1613          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1614          [<code class="option">--enable-stats</code>]
1615        </span></dt><dd><p>Statistics on <code class="varname">arena.&lt;i&gt;.extents_dirty
1616        </code> mutex (arena scope; dirty extents related).
1617        <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1618        counters</a>.</p></dd><dt><a name="stats.arenas.i.mutexes.extents_muzzy"/><span class="term">
1619          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.extents_muzzy.{counter}</code></quote>
1620          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1621          [<code class="option">--enable-stats</code>]
1622        </span></dt><dd><p>Statistics on <code class="varname">arena.&lt;i&gt;.extents_muzzy
1623        </code> mutex (arena scope; muzzy extents related).
1624        <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1625        counters</a>.</p></dd><dt><a name="stats.arenas.i.mutexes.extents_retained"/><span class="term">
1626          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.extents_retained.{counter}</code></quote>
1627          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1628          [<code class="option">--enable-stats</code>]
1629        </span></dt><dd><p>Statistics on <code class="varname">arena.&lt;i&gt;.extents_retained
1630        </code> mutex (arena scope; retained extents related).
1631        <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1632        counters</a>.</p></dd><dt><a name="stats.arenas.i.mutexes.decay_dirty"/><span class="term">
1633          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.decay_dirty.{counter}</code></quote>
1634          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1635          [<code class="option">--enable-stats</code>]
1636        </span></dt><dd><p>Statistics on <code class="varname">arena.&lt;i&gt;.decay_dirty
1637        </code> mutex (arena scope; decay for dirty pages related).
1638        <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1639        counters</a>.</p></dd><dt><a name="stats.arenas.i.mutexes.decay_muzzy"/><span class="term">
1640          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.decay_muzzy.{counter}</code></quote>
1641          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1642          [<code class="option">--enable-stats</code>]
1643        </span></dt><dd><p>Statistics on <code class="varname">arena.&lt;i&gt;.decay_muzzy
1644        </code> mutex (arena scope; decay for muzzy pages related).
1645        <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1646        counters</a>.</p></dd><dt><a name="stats.arenas.i.mutexes.base"/><span class="term">
1647          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.base.{counter}</code></quote>
1648          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1649          [<code class="option">--enable-stats</code>]
1650        </span></dt><dd><p>Statistics on <code class="varname">arena.&lt;i&gt;.base</code>
1651        mutex (arena scope; base allocator related).
1652        <quote><code class="mallctl">{counter}</code></quote> is one of the counters in <a class="link" href="#mutex_counters">mutex profiling
1653        counters</a>.</p></dd><dt><a name="stats.arenas.i.mutexes.tcache_list"/><span class="term">
1654          <quote><code class="mallctl">stats.arenas.&lt;i&gt;.mutexes.tcache_list.{counter}</code></quote>
1655          (<span class="type">counter specific type</span>) <code class="literal">r-</code>
1656          [<code class="option">--enable-stats</code>]
1657        </span></dt><dd><p>Statistics on
1658        <code class="varname">arena.&lt;i&gt;.tcache_list</code> mutex (arena scope;
1659        tcache to arena association related).  This mutex is expected to be
1660        accessed less often.  <quote><code class="mallctl">{counter}</code></quote> is one of the
1661        counters in <a class="link" href="#mutex_counters">mutex profiling
1662        counters</a>.</p></dd></dl></div></div><div class="refsect1"><a name="heap_profile_format"/><h2>HEAP PROFILE FORMAT</h2><p>Although the heap profiling functionality was originally designed to
1663    be compatible with the
1664    <span class="command"><strong>pprof</strong></span> command that is developed as part of the <a class="ulink" href="http://code.google.com/p/gperftools/" target="_top">gperftools
1665    package</a>, the addition of per thread heap profiling functionality
1666    required a different heap profile format.  The <span class="command"><strong>jeprof</strong></span>
1667    command is derived from <span class="command"><strong>pprof</strong></span>, with enhancements to
1668    support the heap profile format described here.</p><p>In the following hypothetical heap profile, <code class="constant">[...]</code>
1669    indicates elision for the sake of compactness.  </p><pre class="programlisting">
1670heap_v2/524288
1671  t*: 28106: 56637512 [0: 0]
1672  [...]
1673  t3: 352: 16777344 [0: 0]
1674  [...]
1675  t99: 17754: 29341640 [0: 0]
1676  [...]
1677@ 0x5f86da8 0x5f5a1dc [...] 0x29e4d4e 0xa200316 0xabb2988 [...]
1678  t*: 13: 6688 [0: 0]
1679  t3: 12: 6496 [0: ]
1680  t99: 1: 192 [0: 0]
1681[...]
1682
1683MAPPED_LIBRARIES:
1684[...]</pre><p> The following matches the above heap profile, but most
1685tokens are replaced with <code class="constant">&lt;description&gt;</code> to indicate
1686descriptions of the corresponding fields.  </p><pre class="programlisting">
1687&lt;heap_profile_format_version&gt;/&lt;mean_sample_interval&gt;
1688  &lt;aggregate&gt;: &lt;curobjs&gt;: &lt;curbytes&gt; [&lt;cumobjs&gt;: &lt;cumbytes&gt;]
1689  [...]
1690  &lt;thread_3_aggregate&gt;: &lt;curobjs&gt;: &lt;curbytes&gt;[&lt;cumobjs&gt;: &lt;cumbytes&gt;]
1691  [...]
1692  &lt;thread_99_aggregate&gt;: &lt;curobjs&gt;: &lt;curbytes&gt;[&lt;cumobjs&gt;: &lt;cumbytes&gt;]
1693  [...]
1694@ &lt;top_frame&gt; &lt;frame&gt; [...] &lt;frame&gt; &lt;frame&gt; &lt;frame&gt; [...]
1695  &lt;backtrace_aggregate&gt;: &lt;curobjs&gt;: &lt;curbytes&gt; [&lt;cumobjs&gt;: &lt;cumbytes&gt;]
1696  &lt;backtrace_thread_3&gt;: &lt;curobjs&gt;: &lt;curbytes&gt; [&lt;cumobjs&gt;: &lt;cumbytes&gt;]
1697  &lt;backtrace_thread_99&gt;: &lt;curobjs&gt;: &lt;curbytes&gt; [&lt;cumobjs&gt;: &lt;cumbytes&gt;]
1698[...]
1699
1700MAPPED_LIBRARIES:
1701&lt;/proc/&lt;pid&gt;/maps&gt;</pre></div><div class="refsect1"><a name="debugging_malloc_problems"/><h2>DEBUGGING MALLOC PROBLEMS</h2><p>When debugging, it is a good idea to configure/build jemalloc with
1702    the <code class="option">--enable-debug</code> and <code class="option">--enable-fill</code>
1703    options, and recompile the program with suitable options and symbols for
1704    debugger support.  When so configured, jemalloc incorporates a wide variety
1705    of run-time assertions that catch application errors such as double-free,
1706    write-after-free, etc.</p><p>Programs often accidentally depend on <span class="quote">“<span class="quote">uninitialized</span>”</span>
1707    memory actually being filled with zero bytes.  Junk filling
1708    (see the <a class="link" href="#opt.junk"><quote><code class="mallctl">opt.junk</code></quote></a>
1709    option) tends to expose such bugs in the form of obviously incorrect
1710    results and/or coredumps.  Conversely, zero
1711    filling (see the <a class="link" href="#opt.zero"><quote><code class="mallctl">opt.zero</code></quote></a> option) eliminates
1712    the symptoms of such bugs.  Between these two options, it is usually
1713    possible to quickly detect, diagnose, and eliminate such bugs.</p><p>This implementation does not provide much detail about the problems
1714    it detects, because the performance impact for storing such information
1715    would be prohibitive.</p></div><div class="refsect1"><a name="diagnostic_messages"/><h2>DIAGNOSTIC MESSAGES</h2><p>If any of the memory allocation/deallocation functions detect an
1716    error or warning condition, a message will be printed to file descriptor
1717    <code class="constant">STDERR_FILENO</code>.  Errors will result in the process
1718    dumping core.  If the <a class="link" href="#opt.abort"><quote><code class="mallctl">opt.abort</code></quote></a> option is set, most
1719    warnings are treated as errors.</p><p>The <code class="varname">malloc_message</code> variable allows the programmer
1720    to override the function which emits the text strings forming the errors
1721    and warnings if for some reason the <code class="constant">STDERR_FILENO</code> file
1722    descriptor is not suitable for this.
1723    <code class="function">malloc_message()</code> takes the
1724    <em class="parameter"><code>cbopaque</code></em> pointer argument that is
1725    <code class="constant">NULL</code> unless overridden by the arguments in a call to
1726    <code class="function">malloc_stats_print()</code>, followed by a string
1727    pointer.  Please note that doing anything which tries to allocate memory in
1728    this function is likely to result in a crash or deadlock.</p><p>All messages are prefixed by
1729    <span class="quote">“<span class="quote"><code class="computeroutput">&lt;jemalloc&gt;: </code></span>”</span>.</p></div><div class="refsect1"><a name="return_values"/><h2>RETURN VALUES</h2><div class="refsect2"><a name="idm140689764475728"/><h3>Standard API</h3><p>The <code class="function">malloc()</code> and
1730      <code class="function">calloc()</code> functions return a pointer to the
1731      allocated memory if successful; otherwise a <code class="constant">NULL</code>
1732      pointer is returned and <code class="varname">errno</code> is set to
1733      <span class="errorname">ENOMEM</span>.</p><p>The <code class="function">posix_memalign()</code> function
1734      returns the value 0 if successful; otherwise it returns an error value.
1735      The <code class="function">posix_memalign()</code> function will fail
1736      if:
1737        </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p>The <em class="parameter"><code>alignment</code></em> parameter is
1738            not a power of 2 at least as large as
1739            <code class="code">sizeof(<span class="type">void *</span>)</code>.
1740            </p></dd><dt><span class="term"><span class="errorname">ENOMEM</span></span></dt><dd><p>Memory allocation error.</p></dd></dl></div><p>
1741      </p><p>The <code class="function">aligned_alloc()</code> function returns
1742      a pointer to the allocated memory if successful; otherwise a
1743      <code class="constant">NULL</code> pointer is returned and
1744      <code class="varname">errno</code> is set.  The
1745      <code class="function">aligned_alloc()</code> function will fail if:
1746        </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p>The <em class="parameter"><code>alignment</code></em> parameter is
1747            not a power of 2.
1748            </p></dd><dt><span class="term"><span class="errorname">ENOMEM</span></span></dt><dd><p>Memory allocation error.</p></dd></dl></div><p>
1749      </p><p>The <code class="function">realloc()</code> function returns a
1750      pointer, possibly identical to <em class="parameter"><code>ptr</code></em>, to the
1751      allocated memory if successful; otherwise a <code class="constant">NULL</code>
1752      pointer is returned, and <code class="varname">errno</code> is set to
1753      <span class="errorname">ENOMEM</span> if the error was the result of an
1754      allocation failure.  The <code class="function">realloc()</code>
1755      function always leaves the original buffer intact when an error occurs.
1756      </p><p>The <code class="function">free()</code> function returns no
1757      value.</p></div><div class="refsect2"><a name="idm140689764454240"/><h3>Non-standard API</h3><p>The <code class="function">mallocx()</code> and
1758      <code class="function">rallocx()</code> functions return a pointer to
1759      the allocated memory if successful; otherwise a <code class="constant">NULL</code>
1760      pointer is returned to indicate insufficient contiguous memory was
1761      available to service the allocation request.  </p><p>The <code class="function">xallocx()</code> function returns the
1762      real size of the resulting resized allocation pointed to by
1763      <em class="parameter"><code>ptr</code></em>, which is a value less than
1764      <em class="parameter"><code>size</code></em> if the allocation could not be adequately
1765      grown in place.  </p><p>The <code class="function">sallocx()</code> function returns the
1766      real size of the allocation pointed to by <em class="parameter"><code>ptr</code></em>.
1767      </p><p>The <code class="function">nallocx()</code> returns the real size
1768      that would result from a successful equivalent
1769      <code class="function">mallocx()</code> function call, or zero if
1770      insufficient memory is available to perform the size computation.  </p><p>The <code class="function">mallctl()</code>,
1771      <code class="function">mallctlnametomib()</code>, and
1772      <code class="function">mallctlbymib()</code> functions return 0 on
1773      success; otherwise they return an error value.  The functions will fail
1774      if:
1775        </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p><em class="parameter"><code>newp</code></em> is not
1776            <code class="constant">NULL</code>, and <em class="parameter"><code>newlen</code></em> is too
1777            large or too small.  Alternatively, <em class="parameter"><code>*oldlenp</code></em>
1778            is too large or too small; in this case as much data as possible
1779            are read despite the error.</p></dd><dt><span class="term"><span class="errorname">ENOENT</span></span></dt><dd><p><em class="parameter"><code>name</code></em> or
1780            <em class="parameter"><code>mib</code></em> specifies an unknown/invalid
1781            value.</p></dd><dt><span class="term"><span class="errorname">EPERM</span></span></dt><dd><p>Attempt to read or write void value, or attempt to
1782            write read-only value.</p></dd><dt><span class="term"><span class="errorname">EAGAIN</span></span></dt><dd><p>A memory allocation failure
1783            occurred.</p></dd><dt><span class="term"><span class="errorname">EFAULT</span></span></dt><dd><p>An interface with side effects failed in some way
1784            not directly related to <code class="function">mallctl*()</code>
1785            read/write processing.</p></dd></dl></div><p>
1786      </p><p>The <code class="function">malloc_usable_size()</code> function
1787      returns the usable size of the allocation pointed to by
1788      <em class="parameter"><code>ptr</code></em>.  </p></div></div><div class="refsect1"><a name="environment"/><h2>ENVIRONMENT</h2><p>The following environment variable affects the execution of the
1789    allocation functions:
1790      </p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><code class="envar">MALLOC_CONF</code></span></dt><dd><p>If the environment variable
1791          <code class="envar">MALLOC_CONF</code> is set, the characters it contains
1792          will be interpreted as options.</p></dd></dl></div><p>
1793    </p></div><div class="refsect1"><a name="examples"/><h2>EXAMPLES</h2><p>To dump core whenever a problem occurs:
1794      </p><pre class="screen">ln -s 'abort:true' /etc/malloc.conf</pre><p>
1795    </p><p>To specify in the source that only one arena should be automatically
1796    created:
1797      </p><pre class="programlisting">
1798malloc_conf = "narenas:1";</pre></div><div class="refsect1"><a name="see_also"/><h2>SEE ALSO</h2><p><span class="citerefentry"><span class="refentrytitle">madvise</span>(2)</span>,
1799    <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span>,
1800    <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>,
1801    <span class="citerefentry"><span class="refentrytitle">utrace</span>(2)</span>,
1802    <span class="citerefentry"><span class="refentrytitle">alloca</span>(3)</span>,
1803    <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span>,
1804    <span class="citerefentry"><span class="refentrytitle">getpagesize</span>(3)</span></p></div><div class="refsect1"><a name="standards"/><h2>STANDARDS</h2><p>The <code class="function">malloc()</code>,
1805    <code class="function">calloc()</code>,
1806    <code class="function">realloc()</code>, and
1807    <code class="function">free()</code> functions conform to ISO/IEC
1808    9899:1990 (<span class="quote">“<span class="quote">ISO C90</span>”</span>).</p><p>The <code class="function">posix_memalign()</code> function conforms
1809    to IEEE Std 1003.1-2001 (<span class="quote">“<span class="quote">POSIX.1</span>”</span>).</p></div></div></body></html>