Sophie

Sophie

distrib > Mandriva > 2010.1 > x86_64 > media > contrib-backports > by-pkgid > 3ba3bd1608c672ba2129b098a48e9e4d > files > 480

python3-docs-3.2.2-3mdv2010.2.noarch.rpm



<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
  "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
    
    <title>Initialization, Finalization, and Threads &mdash; Python v3.2.2 documentation</title>
    <link rel="stylesheet" href="../_static/default.css" type="text/css" />
    <link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
    <script type="text/javascript">
      var DOCUMENTATION_OPTIONS = {
        URL_ROOT:    '../',
        VERSION:     '3.2.2',
        COLLAPSE_INDEX: false,
        FILE_SUFFIX: '.html',
        HAS_SOURCE:  true
      };
    </script>
    <script type="text/javascript" src="../_static/jquery.js"></script>
    <script type="text/javascript" src="../_static/underscore.js"></script>
    <script type="text/javascript" src="../_static/doctools.js"></script>
    <script type="text/javascript" src="../_static/sidebar.js"></script>
    <link rel="search" type="application/opensearchdescription+xml"
          title="Search within Python v3.2.2 documentation"
          href="../_static/opensearch.xml"/>
    <link rel="author" title="About these documents" href="../about.html" />
    <link rel="copyright" title="Copyright" href="../copyright.html" />
    <link rel="top" title="Python v3.2.2 documentation" href="../index.html" />
    <link rel="up" title="Python/C API Reference Manual" href="index.html" />
    <link rel="next" title="Memory Management" href="memory.html" />
    <link rel="prev" title="Code Objects" href="code.html" />
    <link rel="shortcut icon" type="image/png" href="../_static/py.png" />
 

  </head>
  <body>
    <div class="related">
      <h3>Navigation</h3>
      <ul>
        <li class="right" style="margin-right: 10px">
          <a href="../genindex.html" title="General Index"
             accesskey="I">index</a></li>
        <li class="right" >
          <a href="../py-modindex.html" title="Python Module Index"
             >modules</a> |</li>
        <li class="right" >
          <a href="memory.html" title="Memory Management"
             accesskey="N">next</a> |</li>
        <li class="right" >
          <a href="code.html" title="Code Objects"
             accesskey="P">previous</a> |</li>
        <li><img src="../_static/py.png" alt=""
                 style="vertical-align: middle; margin-top: -1px"/></li>
        <li><a href="../index.html">Python v3.2.2 documentation</a> &raquo;</li>

          <li><a href="index.html" accesskey="U">Python/C API Reference Manual</a> &raquo;</li> 
      </ul>
    </div>  

    <div class="document">
      <div class="documentwrapper">
        <div class="bodywrapper">
          <div class="body">
            
  <div class="section" id="initialization-finalization-and-threads">
<span id="initialization"></span><h1>Initialization, Finalization, and Threads<a class="headerlink" href="#initialization-finalization-and-threads" title="Permalink to this headline">¶</a></h1>
<div class="section" id="initializing-and-finalizing-the-interpreter">
<h2>Initializing and finalizing the interpreter<a class="headerlink" href="#initializing-and-finalizing-the-interpreter" title="Permalink to this headline">¶</a></h2>
<dl class="function">
<dt id="Py_Initialize">
void <tt class="descname">Py_Initialize</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_Initialize" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-0">Initialize the Python interpreter.  In an application embedding  Python, this
should be called before using any other Python/C API functions; with the
exception of <a class="reference internal" href="#Py_SetProgramName" title="Py_SetProgramName"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetProgramName()</span></tt></a> and <a class="reference internal" href="#Py_SetPath" title="Py_SetPath"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetPath()</span></tt></a>.  This initializes
the table of loaded modules (<tt class="docutils literal"><span class="pre">sys.modules</span></tt>), and creates the fundamental
modules <a class="reference internal" href="../library/builtins.html#module-builtins" title="builtins: The module that provides the built-in namespace."><tt class="xref py py-mod docutils literal"><span class="pre">builtins</span></tt></a>, <a class="reference internal" href="../library/__main__.html#module-__main__" title="__main__: The environment where the top-level script is run."><tt class="xref py py-mod docutils literal"><span class="pre">__main__</span></tt></a> and <a class="reference internal" href="../library/sys.html#module-sys" title="sys: Access system-specific parameters and functions."><tt class="xref py py-mod docutils literal"><span class="pre">sys</span></tt></a>.  It also initializes
the module search path (<tt class="docutils literal"><span class="pre">sys.path</span></tt>). It does not set <tt class="docutils literal"><span class="pre">sys.argv</span></tt>; use
<a class="reference internal" href="#PySys_SetArgvEx" title="PySys_SetArgvEx"><tt class="xref c c-func docutils literal"><span class="pre">PySys_SetArgvEx()</span></tt></a> for that.  This is a no-op when called for a second time
(without calling <a class="reference internal" href="#Py_Finalize" title="Py_Finalize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Finalize()</span></tt></a> first).  There is no return value; it is a
fatal error if the initialization fails.</p>
</dd></dl>

<dl class="function">
<dt id="Py_InitializeEx">
void <tt class="descname">Py_InitializeEx</tt><big>(</big>int<em>&nbsp;initsigs</em><big>)</big><a class="headerlink" href="#Py_InitializeEx" title="Permalink to this definition">¶</a></dt>
<dd><p>This function works like <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a> if <em>initsigs</em> is 1. If
<em>initsigs</em> is 0, it skips initialization registration of signal handlers, which
might be useful when Python is embedded.</p>
</dd></dl>

<dl class="function">
<dt id="Py_IsInitialized">
int <tt class="descname">Py_IsInitialized</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_IsInitialized" title="Permalink to this definition">¶</a></dt>
<dd><p>Return true (nonzero) when the Python interpreter has been initialized, false
(zero) if not.  After <a class="reference internal" href="#Py_Finalize" title="Py_Finalize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Finalize()</span></tt></a> is called, this returns false until
<a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a> is called again.</p>
</dd></dl>

<dl class="function">
<dt id="Py_Finalize">
void <tt class="descname">Py_Finalize</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_Finalize" title="Permalink to this definition">¶</a></dt>
<dd><p>Undo all initializations made by <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a> and subsequent use of
Python/C API functions, and destroy all sub-interpreters (see
<a class="reference internal" href="#Py_NewInterpreter" title="Py_NewInterpreter"><tt class="xref c c-func docutils literal"><span class="pre">Py_NewInterpreter()</span></tt></a> below) that were created and not yet destroyed since
the last call to <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a>.  Ideally, this frees all memory
allocated by the Python interpreter.  This is a no-op when called for a second
time (without calling <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a> again first).  There is no return
value; errors during finalization are ignored.</p>
<p>This function is provided for a number of reasons.  An embedding application
might want to restart Python without having to restart the application itself.
An application that has loaded the Python interpreter from a dynamically
loadable library (or DLL) might want to free all memory allocated by Python
before unloading the DLL. During a hunt for memory leaks in an application a
developer might want to free all memory allocated by Python before exiting from
the application.</p>
<p><strong>Bugs and caveats:</strong> The destruction of modules and objects in modules is done
in random order; this may cause destructors (<a class="reference internal" href="../reference/datamodel.html#object.__del__" title="object.__del__"><tt class="xref py py-meth docutils literal"><span class="pre">__del__()</span></tt></a> methods) to fail
when they depend on other objects (even functions) or modules.  Dynamically
loaded extension modules loaded by Python are not unloaded.  Small amounts of
memory allocated by the Python interpreter may not be freed (if you find a leak,
please report it).  Memory tied up in circular references between objects is not
freed.  Some memory allocated by extension modules may not be freed.  Some
extensions may not work properly if their initialization routine is called more
than once; this can happen if an application calls <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a> and
<a class="reference internal" href="#Py_Finalize" title="Py_Finalize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Finalize()</span></tt></a> more than once.</p>
</dd></dl>

</div>
<div class="section" id="process-wide-parameters">
<h2>Process-wide parameters<a class="headerlink" href="#process-wide-parameters" title="Permalink to this headline">¶</a></h2>
<dl class="function">
<dt id="Py_SetProgramName">
void <tt class="descname">Py_SetProgramName</tt><big>(</big>wchar_t<em>&nbsp;*name</em><big>)</big><a class="headerlink" href="#Py_SetProgramName" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-1">This function should be called before <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a> is called for
the first time, if it is called at all.  It tells the interpreter the value
of the <tt class="docutils literal"><span class="pre">argv[0]</span></tt> argument to the <tt class="xref c c-func docutils literal"><span class="pre">main()</span></tt> function of the program
(converted to wide characters).
This is used by <a class="reference internal" href="#Py_GetPath" title="Py_GetPath"><tt class="xref c c-func docutils literal"><span class="pre">Py_GetPath()</span></tt></a> and some other functions below to find
the Python run-time libraries relative to the interpreter executable.  The
default value is <tt class="docutils literal"><span class="pre">'python'</span></tt>.  The argument should point to a
zero-terminated wide character string in static storage whose contents will not
change for the duration of the program&#8217;s execution.  No code in the Python
interpreter will change the contents of this storage.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetProgramName">
wchar* <tt class="descname">Py_GetProgramName</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetProgramName" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-2">Return the program name set with <a class="reference internal" href="#Py_SetProgramName" title="Py_SetProgramName"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetProgramName()</span></tt></a>, or the default.
The returned string points into static storage; the caller should not modify its
value.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetPrefix">
wchar_t* <tt class="descname">Py_GetPrefix</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetPrefix" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the <em>prefix</em> for installed platform-independent files. This is derived
through a number of complicated rules from the program name set with
<a class="reference internal" href="#Py_SetProgramName" title="Py_SetProgramName"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetProgramName()</span></tt></a> and some environment variables; for example, if the
program name is <tt class="docutils literal"><span class="pre">'/usr/bin/python3'</span></tt>, the prefix is <tt class="docutils literal"><span class="pre">'/usr/local'</span></tt>. The
returned string points into static storage; the caller should not modify its
value.  This corresponds to the <strong class="makevar">prefix</strong> variable in the top-level
<tt class="file docutils literal"><span class="pre">Makefile</span></tt> and the <tt class="docutils literal"><span class="pre">--prefix</span></tt> argument to the <strong class="program">configure</strong>
script at build time.  The value is available to Python code as <tt class="docutils literal"><span class="pre">sys.prefix</span></tt>.
It is only useful on Unix.  See also the next function.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetExecPrefix">
wchar_t* <tt class="descname">Py_GetExecPrefix</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetExecPrefix" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the <em>exec-prefix</em> for installed platform-<em>dependent</em> files.  This is
derived through a number of complicated rules from the program name set with
<a class="reference internal" href="#Py_SetProgramName" title="Py_SetProgramName"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetProgramName()</span></tt></a> and some environment variables; for example, if the
program name is <tt class="docutils literal"><span class="pre">'/usr/bin/python3'</span></tt>, the exec-prefix is
<tt class="docutils literal"><span class="pre">'/usr/local'</span></tt>.  The returned string points into static storage; the caller
should not modify its value.  This corresponds to the <strong class="makevar">exec_prefix</strong>
variable in the top-level <tt class="file docutils literal"><span class="pre">Makefile</span></tt> and the <tt class="docutils literal"><span class="pre">--exec-prefix</span></tt>
argument to the <strong class="program">configure</strong> script at build  time.  The value is
available to Python code as <tt class="docutils literal"><span class="pre">sys.exec_prefix</span></tt>.  It is only useful on Unix.</p>
<p>Background: The exec-prefix differs from the prefix when platform dependent
files (such as executables and shared libraries) are installed in a different
directory tree.  In a typical installation, platform dependent files may be
installed in the <tt class="file docutils literal"><span class="pre">/usr/local/plat</span></tt> subtree while platform independent may
be installed in <tt class="file docutils literal"><span class="pre">/usr/local</span></tt>.</p>
<p>Generally speaking, a platform is a combination of hardware and software
families, e.g.  Sparc machines running the Solaris 2.x operating system are
considered the same platform, but Intel machines running Solaris 2.x are another
platform, and Intel machines running Linux are yet another platform.  Different
major revisions of the same operating system generally also form different
platforms.  Non-Unix operating systems are a different story; the installation
strategies on those systems are so different that the prefix and exec-prefix are
meaningless, and set to the empty string. Note that compiled Python bytecode
files are platform independent (but not independent from the Python version by
which they were compiled!).</p>
<p>System administrators will know how to configure the <strong class="program">mount</strong> or
<strong class="program">automount</strong> programs to share <tt class="file docutils literal"><span class="pre">/usr/local</span></tt> between platforms
while having <tt class="file docutils literal"><span class="pre">/usr/local/plat</span></tt> be a different filesystem for each
platform.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetProgramFullPath">
wchar_t* <tt class="descname">Py_GetProgramFullPath</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetProgramFullPath" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-3">Return the full program name of the Python executable; this is  computed as a
side-effect of deriving the default module search path  from the program name
(set by <a class="reference internal" href="#Py_SetProgramName" title="Py_SetProgramName"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetProgramName()</span></tt></a> above). The returned string points into
static storage; the caller should not modify its value.  The value is available
to Python code as <tt class="docutils literal"><span class="pre">sys.executable</span></tt>.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetPath">
wchar_t* <tt class="descname">Py_GetPath</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetPath" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-4">Return the default module search path; this is computed from the program name
(set by <a class="reference internal" href="#Py_SetProgramName" title="Py_SetProgramName"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetProgramName()</span></tt></a> above) and some environment variables.
The returned string consists of a series of directory names separated by a
platform dependent delimiter character.  The delimiter character is <tt class="docutils literal"><span class="pre">':'</span></tt>
on Unix and Mac OS X, <tt class="docutils literal"><span class="pre">';'</span></tt> on Windows.  The returned string points into
static storage; the caller should not modify its value.  The list
<a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><tt class="xref py py-data docutils literal"><span class="pre">sys.path</span></tt></a> is initialized with this value on interpreter startup; it
can be (and usually is) modified later to change the search path for loading
modules.</p>
</dd></dl>

<dl class="function">
<dt id="Py_SetPath">
void <tt class="descname">Py_SetPath</tt><big>(</big>const wchar_t<em>&nbsp;*</em><big>)</big><a class="headerlink" href="#Py_SetPath" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-5">Set the default module search path.  If this function is called before
<a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a>, then <a class="reference internal" href="#Py_GetPath" title="Py_GetPath"><tt class="xref c c-func docutils literal"><span class="pre">Py_GetPath()</span></tt></a> won&#8217;t attempt to compute a
default search path but uses the one provided instead.  This is useful if
Python is embedded by an application that has full knowledge of the location
of all modules.  The path components should be separated by semicolons.</p>
<p>This also causes <a class="reference internal" href="../library/sys.html#sys.executable" title="sys.executable"><tt class="xref py py-data docutils literal"><span class="pre">sys.executable</span></tt></a> to be set only to the raw program
name (see <a class="reference internal" href="#Py_SetProgramName" title="Py_SetProgramName"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetProgramName()</span></tt></a>) and for <a class="reference internal" href="../library/sys.html#sys.prefix" title="sys.prefix"><tt class="xref py py-data docutils literal"><span class="pre">sys.prefix</span></tt></a> and
<a class="reference internal" href="../library/sys.html#sys.exec_prefix" title="sys.exec_prefix"><tt class="xref py py-data docutils literal"><span class="pre">sys.exec_prefix</span></tt></a> to be empty.  It is up to the caller to modify these
if required after calling <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a>.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetVersion">
const char* <tt class="descname">Py_GetVersion</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetVersion" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the version of this Python interpreter.  This is a string that looks
something like</p>
<div class="highlight-c"><div class="highlight"><pre><span class="s">&quot;3.0a5+ (py3k:63103M, May 12 2008, 00:53:55) </span><span class="se">\n</span><span class="s">[GCC 4.2.3]&quot;</span>
</pre></div>
</div>
<p id="index-6">The first word (up to the first space character) is the current Python version;
the first three characters are the major and minor version separated by a
period.  The returned string points into static storage; the caller should not
modify its value.  The value is available to Python code as <a class="reference internal" href="../library/sys.html#sys.version" title="sys.version"><tt class="xref py py-data docutils literal"><span class="pre">sys.version</span></tt></a>.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetPlatform">
const char* <tt class="descname">Py_GetPlatform</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetPlatform" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-7">Return the platform identifier for the current platform.  On Unix, this is
formed from the &#8220;official&#8221; name of the operating system, converted to lower
case, followed by the major revision number; e.g., for Solaris 2.x, which is
also known as SunOS 5.x, the value is <tt class="docutils literal"><span class="pre">'sunos5'</span></tt>.  On Mac OS X, it is
<tt class="docutils literal"><span class="pre">'darwin'</span></tt>.  On Windows, it is <tt class="docutils literal"><span class="pre">'win'</span></tt>.  The returned string points into
static storage; the caller should not modify its value.  The value is available
to Python code as <tt class="docutils literal"><span class="pre">sys.platform</span></tt>.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetCopyright">
const char* <tt class="descname">Py_GetCopyright</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetCopyright" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the official copyright string for the current Python version, for example</p>
<p><tt class="docutils literal"><span class="pre">'Copyright</span> <span class="pre">1991-1995</span> <span class="pre">Stichting</span> <span class="pre">Mathematisch</span> <span class="pre">Centrum,</span> <span class="pre">Amsterdam'</span></tt></p>
<p id="index-8">The returned string points into static storage; the caller should not modify its
value.  The value is available to Python code as <tt class="docutils literal"><span class="pre">sys.copyright</span></tt>.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetCompiler">
const char* <tt class="descname">Py_GetCompiler</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetCompiler" title="Permalink to this definition">¶</a></dt>
<dd><p>Return an indication of the compiler used to build the current Python version,
in square brackets, for example:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="s">&quot;[GCC 2.7.2.2]&quot;</span>
</pre></div>
</div>
<p id="index-9">The returned string points into static storage; the caller should not modify its
value.  The value is available to Python code as part of the variable
<tt class="docutils literal"><span class="pre">sys.version</span></tt>.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetBuildInfo">
const char* <tt class="descname">Py_GetBuildInfo</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetBuildInfo" title="Permalink to this definition">¶</a></dt>
<dd><p>Return information about the sequence number and build date and time  of the
current Python interpreter instance, for example</p>
<div class="highlight-c"><div class="highlight"><pre><span class="s">&quot;#67, Aug  1 1997, 22:34:28&quot;</span>
</pre></div>
</div>
<p id="index-10">The returned string points into static storage; the caller should not modify its
value.  The value is available to Python code as part of the variable
<tt class="docutils literal"><span class="pre">sys.version</span></tt>.</p>
</dd></dl>

<dl class="function">
<dt id="PySys_SetArgvEx">
void <tt class="descname">PySys_SetArgvEx</tt><big>(</big>int<em>&nbsp;argc</em>, wchar_t<em>&nbsp;**argv</em>, int<em>&nbsp;updatepath</em><big>)</big><a class="headerlink" href="#PySys_SetArgvEx" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-11">Set <a class="reference internal" href="../library/sys.html#sys.argv" title="sys.argv"><tt class="xref py py-data docutils literal"><span class="pre">sys.argv</span></tt></a> based on <em>argc</em> and <em>argv</em>.  These parameters are
similar to those passed to the program&#8217;s <tt class="xref c c-func docutils literal"><span class="pre">main()</span></tt> function with the
difference that the first entry should refer to the script file to be
executed rather than the executable hosting the Python interpreter.  If there
isn&#8217;t a script that will be run, the first entry in <em>argv</em> can be an empty
string.  If this function fails to initialize <a class="reference internal" href="../library/sys.html#sys.argv" title="sys.argv"><tt class="xref py py-data docutils literal"><span class="pre">sys.argv</span></tt></a>, a fatal
condition is signalled using <a class="reference internal" href="sys.html#Py_FatalError" title="Py_FatalError"><tt class="xref c c-func docutils literal"><span class="pre">Py_FatalError()</span></tt></a>.</p>
<p>If <em>updatepath</em> is zero, this is all the function does.  If <em>updatepath</em>
is non-zero, the function also modifies <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><tt class="xref py py-data docutils literal"><span class="pre">sys.path</span></tt></a> according to the
following algorithm:</p>
<ul class="simple">
<li>If the name of an existing script is passed in <tt class="docutils literal"><span class="pre">argv[0]</span></tt>, the absolute
path of the directory where the script is located is prepended to
<a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><tt class="xref py py-data docutils literal"><span class="pre">sys.path</span></tt></a>.</li>
<li>Otherwise (that is, if <em>argc</em> is 0 or <tt class="docutils literal"><span class="pre">argv[0]</span></tt> doesn&#8217;t point
to an existing file name), an empty string is prepended to
<a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><tt class="xref py py-data docutils literal"><span class="pre">sys.path</span></tt></a>, which is the same as prepending the current working
directory (<tt class="docutils literal"><span class="pre">&quot;.&quot;</span></tt>).</li>
</ul>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p>It is recommended that applications embedding the Python interpreter
for purposes other than executing a single script pass 0 as <em>updatepath</em>,
and update <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><tt class="xref py py-data docutils literal"><span class="pre">sys.path</span></tt></a> themselves if desired.
See <a class="reference external" href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2008-5983">CVE-2008-5983</a>.</p>
<p>On versions before 3.1.3, you can achieve the same effect by manually
popping the first <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><tt class="xref py py-data docutils literal"><span class="pre">sys.path</span></tt></a> element after having called
<a class="reference internal" href="#PySys_SetArgv" title="PySys_SetArgv"><tt class="xref c c-func docutils literal"><span class="pre">PySys_SetArgv()</span></tt></a>, for example using:</p>
<div class="last highlight-c"><div class="highlight"><pre><span class="n">PyRun_SimpleString</span><span class="p">(</span><span class="s">&quot;import sys; sys.path.pop(0)</span><span class="se">\n</span><span class="s">&quot;</span><span class="p">);</span>
</pre></div>
</div>
</div>
<p class="versionadded">
<span class="versionmodified">New in version 3.1.3.</span></p>
</dd></dl>

<dl class="function">
<dt id="PySys_SetArgv">
void <tt class="descname">PySys_SetArgv</tt><big>(</big>int<em>&nbsp;argc</em>, wchar_t<em>&nbsp;**argv</em><big>)</big><a class="headerlink" href="#PySys_SetArgv" title="Permalink to this definition">¶</a></dt>
<dd><p>This function works like <a class="reference internal" href="#PySys_SetArgvEx" title="PySys_SetArgvEx"><tt class="xref c c-func docutils literal"><span class="pre">PySys_SetArgvEx()</span></tt></a> with <em>updatepath</em> set to 1.</p>
</dd></dl>

<dl class="function">
<dt id="Py_SetPythonHome">
void <tt class="descname">Py_SetPythonHome</tt><big>(</big>wchar_t<em>&nbsp;*home</em><big>)</big><a class="headerlink" href="#Py_SetPythonHome" title="Permalink to this definition">¶</a></dt>
<dd><p>Set the default &#8220;home&#8221; directory, that is, the location of the standard
Python libraries.  See <span class="target" id="index-12"></span><a class="reference internal" href="../using/cmdline.html#envvar-PYTHONHOME"><tt class="xref std std-envvar docutils literal"><span class="pre">PYTHONHOME</span></tt></a> for the meaning of the
argument string.</p>
<p>The argument should point to a zero-terminated character string in static
storage whose contents will not change for the duration of the program&#8217;s
execution.  No code in the Python interpreter will change the contents of
this storage.</p>
</dd></dl>

<dl class="function">
<dt id="Py_GetPythonHome">
w_char* <tt class="descname">Py_GetPythonHome</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_GetPythonHome" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the default &#8220;home&#8221;, that is, the value set by a previous call to
<a class="reference internal" href="#Py_SetPythonHome" title="Py_SetPythonHome"><tt class="xref c c-func docutils literal"><span class="pre">Py_SetPythonHome()</span></tt></a>, or the value of the <span class="target" id="index-13"></span><a class="reference internal" href="../using/cmdline.html#envvar-PYTHONHOME"><tt class="xref std std-envvar docutils literal"><span class="pre">PYTHONHOME</span></tt></a>
environment variable if it is set.</p>
</dd></dl>

</div>
<div class="section" id="thread-state-and-the-global-interpreter-lock">
<span id="threads"></span><h2>Thread State and the Global Interpreter Lock<a class="headerlink" href="#thread-state-and-the-global-interpreter-lock" title="Permalink to this headline">¶</a></h2>
<p id="index-14">The Python interpreter is not fully thread-safe.  In order to support
multi-threaded Python programs, there&#8217;s a global lock, called the <a class="reference internal" href="../glossary.html#term-global-interpreter-lock"><em class="xref std std-term">global
interpreter lock</em></a> or <a class="reference internal" href="../glossary.html#term-gil"><em class="xref std std-term">GIL</em></a>, that must be held by the current thread before
it can safely access Python objects. Without the lock, even the simplest
operations could cause problems in a multi-threaded program: for example, when
two threads simultaneously increment the reference count of the same object, the
reference count could end up being incremented only once instead of twice.</p>
<p id="index-15">Therefore, the rule exists that only the thread that has acquired the
<a class="reference internal" href="../glossary.html#term-gil"><em class="xref std std-term">GIL</em></a> may operate on Python objects or call Python/C API functions.
In order to emulate concurrency of execution, the interpreter regularly
tries to switch threads (see <a class="reference internal" href="../library/sys.html#sys.setswitchinterval" title="sys.setswitchinterval"><tt class="xref py py-func docutils literal"><span class="pre">sys.setswitchinterval()</span></tt></a>).  The lock is also
released around potentially blocking I/O operations like reading or writing
a file, so that other Python threads can run in the meantime.</p>
<p id="index-16">The Python interpreter keeps some thread-specific bookkeeping information
inside a data structure called <a class="reference internal" href="#PyThreadState" title="PyThreadState"><tt class="xref c c-type docutils literal"><span class="pre">PyThreadState</span></tt></a>.  There&#8217;s also one
global variable pointing to the current <a class="reference internal" href="#PyThreadState" title="PyThreadState"><tt class="xref c c-type docutils literal"><span class="pre">PyThreadState</span></tt></a>: it can
be retrieved using <a class="reference internal" href="#PyThreadState_Get" title="PyThreadState_Get"><tt class="xref c c-func docutils literal"><span class="pre">PyThreadState_Get()</span></tt></a>.</p>
<div class="section" id="releasing-the-gil-from-extension-code">
<h3>Releasing the GIL from extension code<a class="headerlink" href="#releasing-the-gil-from-extension-code" title="Permalink to this headline">¶</a></h3>
<p>Most extension code manipulating the <a class="reference internal" href="../glossary.html#term-gil"><em class="xref std std-term">GIL</em></a> has the following simple
structure:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="n">Save</span> <span class="n">the</span> <span class="kr">thread</span> <span class="n">state</span> <span class="n">in</span> <span class="n">a</span> <span class="n">local</span> <span class="n">variable</span><span class="p">.</span>
<span class="n">Release</span> <span class="n">the</span> <span class="n">global</span> <span class="n">interpreter</span> <span class="n">lock</span><span class="p">.</span>
<span class="p">...</span> <span class="n">Do</span> <span class="n">some</span> <span class="n">blocking</span> <span class="n">I</span><span class="o">/</span><span class="n">O</span> <span class="n">operation</span> <span class="p">...</span>
<span class="n">Reacquire</span> <span class="n">the</span> <span class="n">global</span> <span class="n">interpreter</span> <span class="n">lock</span><span class="p">.</span>
<span class="n">Restore</span> <span class="n">the</span> <span class="kr">thread</span> <span class="n">state</span> <span class="n">from</span> <span class="n">the</span> <span class="n">local</span> <span class="n">variable</span><span class="p">.</span>
</pre></div>
</div>
<p>This is so common that a pair of macros exists to simplify it:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="n">Py_BEGIN_ALLOW_THREADS</span>
<span class="p">...</span> <span class="n">Do</span> <span class="n">some</span> <span class="n">blocking</span> <span class="n">I</span><span class="o">/</span><span class="n">O</span> <span class="n">operation</span> <span class="p">...</span>
<span class="n">Py_END_ALLOW_THREADS</span>
</pre></div>
</div>
<p id="index-17">The <a class="reference internal" href="#Py_BEGIN_ALLOW_THREADS" title="Py_BEGIN_ALLOW_THREADS"><tt class="xref c c-macro docutils literal"><span class="pre">Py_BEGIN_ALLOW_THREADS</span></tt></a> macro opens a new block and declares a
hidden local variable; the <a class="reference internal" href="#Py_END_ALLOW_THREADS" title="Py_END_ALLOW_THREADS"><tt class="xref c c-macro docutils literal"><span class="pre">Py_END_ALLOW_THREADS</span></tt></a> macro closes the
block.  These two macros are still available when Python is compiled without
thread support (they simply have an empty expansion).</p>
<p>When thread support is enabled, the block above expands to the following code:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="n">PyThreadState</span> <span class="o">*</span><span class="n">_save</span><span class="p">;</span>

<span class="n">_save</span> <span class="o">=</span> <span class="n">PyEval_SaveThread</span><span class="p">();</span>
<span class="p">...</span><span class="n">Do</span> <span class="n">some</span> <span class="n">blocking</span> <span class="n">I</span><span class="o">/</span><span class="n">O</span> <span class="n">operation</span><span class="p">...</span>
<span class="n">PyEval_RestoreThread</span><span class="p">(</span><span class="n">_save</span><span class="p">);</span>
</pre></div>
</div>
<p id="index-18">Here is how these functions work: the global interpreter lock is used to protect the pointer to the
current thread state.  When releasing the lock and saving the thread state,
the current thread state pointer must be retrieved before the lock is released
(since another thread could immediately acquire the lock and store its own thread
state in the global variable). Conversely, when acquiring the lock and restoring
the thread state, the lock must be acquired before storing the thread state
pointer.</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">Calling system I/O functions is the most common use case for releasing
the GIL, but it can also be useful before calling long-running computations
which don&#8217;t need access to Python objects, such as compression or
cryptographic functions operating over memory buffers.  For example, the
standard <a class="reference internal" href="../library/zlib.html#module-zlib" title="zlib: Low-level interface to compression and decompression routines compatible with gzip."><tt class="xref py py-mod docutils literal"><span class="pre">zlib</span></tt></a> and <a class="reference internal" href="../library/hashlib.html#module-hashlib" title="hashlib: Secure hash and message digest algorithms."><tt class="xref py py-mod docutils literal"><span class="pre">hashlib</span></tt></a> modules release the GIL when
compressing or hashing data.</p>
</div>
</div>
<div class="section" id="non-python-created-threads">
<h3>Non-Python created threads<a class="headerlink" href="#non-python-created-threads" title="Permalink to this headline">¶</a></h3>
<p>When threads are created using the dedicated Python APIs (such as the
<a class="reference internal" href="../library/threading.html#module-threading" title="threading: Thread-based parallelism."><tt class="xref py py-mod docutils literal"><span class="pre">threading</span></tt></a> module), a thread state is automatically associated to them
and the code showed above is therefore correct.  However, when threads are
created from C (for example by a third-party library with its own thread
management), they don&#8217;t hold the GIL, nor is there a thread state structure
for them.</p>
<p>If you need to call Python code from these threads (often this will be part
of a callback API provided by the aforementioned third-party library),
you must first register these threads with the interpreter by
creating a thread state data structure, then acquiring the GIL, and finally
storing their thread state pointer, before you can start using the Python/C
API.  When you are done, you should reset the thread state pointer, release
the GIL, and finally free the thread state data structure.</p>
<p>The <a class="reference internal" href="#PyGILState_Ensure" title="PyGILState_Ensure"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Ensure()</span></tt></a> and <a class="reference internal" href="#PyGILState_Release" title="PyGILState_Release"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Release()</span></tt></a> functions do
all of the above automatically.  The typical idiom for calling into Python
from a C thread is:</p>
<div class="highlight-c"><div class="highlight"><pre><span class="n">PyGILState_STATE</span> <span class="n">gstate</span><span class="p">;</span>
<span class="n">gstate</span> <span class="o">=</span> <span class="n">PyGILState_Ensure</span><span class="p">();</span>

<span class="cm">/* Perform Python actions here. */</span>
<span class="n">result</span> <span class="o">=</span> <span class="n">CallSomeFunction</span><span class="p">();</span>
<span class="cm">/* evaluate result or handle exception */</span>

<span class="cm">/* Release the thread. No Python API allowed beyond this point. */</span>
<span class="n">PyGILState_Release</span><span class="p">(</span><span class="n">gstate</span><span class="p">);</span>
</pre></div>
</div>
<p>Note that the <tt class="xref c c-func docutils literal"><span class="pre">PyGILState_*()</span></tt> functions assume there is only one global
interpreter (created automatically by <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a>).  Python
supports the creation of additional interpreters (using
<a class="reference internal" href="#Py_NewInterpreter" title="Py_NewInterpreter"><tt class="xref c c-func docutils literal"><span class="pre">Py_NewInterpreter()</span></tt></a>), but mixing multiple interpreters and the
<tt class="xref c c-func docutils literal"><span class="pre">PyGILState_*()</span></tt> API is unsupported.</p>
<p>Another important thing to note about threads is their behaviour in the face
of the C <tt class="xref c c-func docutils literal"><span class="pre">fork()</span></tt> call. On most systems with <tt class="xref c c-func docutils literal"><span class="pre">fork()</span></tt>, after a
process forks only the thread that issued the fork will exist. That also
means any locks held by other threads will never be released. Python solves
this for <a class="reference internal" href="../library/os.html#os.fork" title="os.fork"><tt class="xref py py-func docutils literal"><span class="pre">os.fork()</span></tt></a> by acquiring the locks it uses internally before
the fork, and releasing them afterwards. In addition, it resets any
<a class="reference internal" href="../library/threading.html#lock-objects"><em>Lock Objects</em></a> in the child. When extending or embedding Python, there
is no way to inform Python of additional (non-Python) locks that need to be
acquired before or reset after a fork. OS facilities such as
<tt class="xref c c-func docutils literal"><span class="pre">pthread_atfork()</span></tt> would need to be used to accomplish the same thing.
Additionally, when extending or embedding Python, calling <tt class="xref c c-func docutils literal"><span class="pre">fork()</span></tt>
directly rather than through <a class="reference internal" href="../library/os.html#os.fork" title="os.fork"><tt class="xref py py-func docutils literal"><span class="pre">os.fork()</span></tt></a> (and returning to or calling
into Python) may result in a deadlock by one of Python&#8217;s internal locks
being held by a thread that is defunct after the fork.
<a class="reference internal" href="sys.html#PyOS_AfterFork" title="PyOS_AfterFork"><tt class="xref c c-func docutils literal"><span class="pre">PyOS_AfterFork()</span></tt></a> tries to reset the necessary locks, but is not
always able to.</p>
</div>
<div class="section" id="high-level-api">
<h3>High-level API<a class="headerlink" href="#high-level-api" title="Permalink to this headline">¶</a></h3>
<p>These are the most commonly used types and functions when writing C extension
code, or when embedding the Python interpreter:</p>
<dl class="type">
<dt id="PyInterpreterState">
<tt class="descname">PyInterpreterState</tt><a class="headerlink" href="#PyInterpreterState" title="Permalink to this definition">¶</a></dt>
<dd><p>This data structure represents the state shared by a number of cooperating
threads.  Threads belonging to the same interpreter share their module
administration and a few other internal items. There are no public members in
this structure.</p>
<p>Threads belonging to different interpreters initially share nothing, except
process state like available memory, open file descriptors and such.  The global
interpreter lock is also shared by all threads, regardless of to which
interpreter they belong.</p>
</dd></dl>

<dl class="type">
<dt id="PyThreadState">
<tt class="descname">PyThreadState</tt><a class="headerlink" href="#PyThreadState" title="Permalink to this definition">¶</a></dt>
<dd><p>This data structure represents the state of a single thread.  The only public
data member is <a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState"><tt class="xref c c-type docutils literal"><span class="pre">PyInterpreterState</span> <span class="pre">*</span></tt></a><tt class="xref py py-attr docutils literal"><span class="pre">interp</span></tt>, which points to
this thread&#8217;s interpreter state.</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_InitThreads">
void <tt class="descname">PyEval_InitThreads</tt><big>(</big><big>)</big><a class="headerlink" href="#PyEval_InitThreads" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-19">Initialize and acquire the global interpreter lock.  It should be called in the
main thread before creating a second thread or engaging in any other thread
operations such as <tt class="docutils literal"><span class="pre">PyEval_ReleaseThread(tstate)</span></tt>. It is not needed before
calling <a class="reference internal" href="#PyEval_SaveThread" title="PyEval_SaveThread"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_SaveThread()</span></tt></a> or <a class="reference internal" href="#PyEval_RestoreThread" title="PyEval_RestoreThread"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_RestoreThread()</span></tt></a>.</p>
<p>This is a no-op when called for a second time.</p>
<p class="versionchanged">
<span class="versionmodified">Changed in version 3.2: </span>This function cannot be called before <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a> anymore.</p>
<div class="admonition note" id="index-20">
<p class="first admonition-title">Note</p>
<p>When only the main thread exists, no GIL operations are needed. This is a
common situation (most Python programs do not use threads), and the lock
operations slow the interpreter down a bit. Therefore, the lock is not
created initially.  This situation is equivalent to having acquired the lock:
when there is only a single thread, all object accesses are safe.  Therefore,
when this function initializes the global interpreter lock, it also acquires
it.  Before the Python <a class="reference internal" href="../library/_thread.html#module-_thread" title="_thread: Low-level threading API."><tt class="xref py py-mod docutils literal"><span class="pre">_thread</span></tt></a> module creates a new thread, knowing
that either it has the lock or the lock hasn&#8217;t been created yet, it calls
<a class="reference internal" href="#PyEval_InitThreads" title="PyEval_InitThreads"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_InitThreads()</span></tt></a>.  When this call returns, it is guaranteed that
the lock has been created and that the calling thread has acquired it.</p>
<p>It is <strong>not</strong> safe to call this function when it is unknown which thread (if
any) currently has the global interpreter lock.</p>
<p class="last">This function is not available when thread support is disabled at compile time.</p>
</div>
</dd></dl>

<dl class="function">
<dt id="PyEval_ThreadsInitialized">
int <tt class="descname">PyEval_ThreadsInitialized</tt><big>(</big><big>)</big><a class="headerlink" href="#PyEval_ThreadsInitialized" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns a non-zero value if <a class="reference internal" href="#PyEval_InitThreads" title="PyEval_InitThreads"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_InitThreads()</span></tt></a> has been called.  This
function can be called without holding the GIL, and therefore can be used to
avoid calls to the locking API when running single-threaded.  This function is
not available when thread support is disabled at compile time.</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_SaveThread">
<a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a>* <tt class="descname">PyEval_SaveThread</tt><big>(</big><big>)</big><a class="headerlink" href="#PyEval_SaveThread" title="Permalink to this definition">¶</a></dt>
<dd><p>Release the global interpreter lock (if it has been created and thread
support is enabled) and reset the thread state to <em>NULL</em>, returning the
previous thread state (which is not <em>NULL</em>).  If the lock has been created,
the current thread must have acquired it.  (This function is available even
when thread support is disabled at compile time.)</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_RestoreThread">
void <tt class="descname">PyEval_RestoreThread</tt><big>(</big><a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a><em>&nbsp;*tstate</em><big>)</big><a class="headerlink" href="#PyEval_RestoreThread" title="Permalink to this definition">¶</a></dt>
<dd><p>Acquire the global interpreter lock (if it has been created and thread
support is enabled) and set the thread state to <em>tstate</em>, which must not be
<em>NULL</em>.  If the lock has been created, the current thread must not have
acquired it, otherwise deadlock ensues.  (This function is available even
when thread support is disabled at compile time.)</p>
</dd></dl>

<dl class="function">
<dt id="PyThreadState_Get">
<a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a>* <tt class="descname">PyThreadState_Get</tt><big>(</big><big>)</big><a class="headerlink" href="#PyThreadState_Get" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the current thread state.  The global interpreter lock must be held.
When the current thread state is <em>NULL</em>, this issues a fatal error (so that
the caller needn&#8217;t check for <em>NULL</em>).</p>
</dd></dl>

<dl class="function">
<dt id="PyThreadState_Swap">
<a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a>* <tt class="descname">PyThreadState_Swap</tt><big>(</big><a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a><em>&nbsp;*tstate</em><big>)</big><a class="headerlink" href="#PyThreadState_Swap" title="Permalink to this definition">¶</a></dt>
<dd><p>Swap the current thread state with the thread state given by the argument
<em>tstate</em>, which may be <em>NULL</em>.  The global interpreter lock must be held
and is not released.</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_ReInitThreads">
void <tt class="descname">PyEval_ReInitThreads</tt><big>(</big><big>)</big><a class="headerlink" href="#PyEval_ReInitThreads" title="Permalink to this definition">¶</a></dt>
<dd><p>This function is called from <a class="reference internal" href="sys.html#PyOS_AfterFork" title="PyOS_AfterFork"><tt class="xref c c-func docutils literal"><span class="pre">PyOS_AfterFork()</span></tt></a> to ensure that newly
created child processes don&#8217;t hold locks referring to threads which
are not running in the child process.</p>
</dd></dl>

<p>The following functions use thread-local storage, and are not compatible
with sub-interpreters:</p>
<dl class="function">
<dt id="PyGILState_Ensure">
PyGILState_STATE <tt class="descname">PyGILState_Ensure</tt><big>(</big><big>)</big><a class="headerlink" href="#PyGILState_Ensure" title="Permalink to this definition">¶</a></dt>
<dd><p>Ensure that the current thread is ready to call the Python C API regardless
of the current state of Python, or of the global interpreter lock. This may
be called as many times as desired by a thread as long as each call is
matched with a call to <a class="reference internal" href="#PyGILState_Release" title="PyGILState_Release"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Release()</span></tt></a>. In general, other
thread-related APIs may be used between <a class="reference internal" href="#PyGILState_Ensure" title="PyGILState_Ensure"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Ensure()</span></tt></a> and
<a class="reference internal" href="#PyGILState_Release" title="PyGILState_Release"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Release()</span></tt></a> calls as long as the thread state is restored to
its previous state before the Release().  For example, normal usage of the
<a class="reference internal" href="#Py_BEGIN_ALLOW_THREADS" title="Py_BEGIN_ALLOW_THREADS"><tt class="xref c c-macro docutils literal"><span class="pre">Py_BEGIN_ALLOW_THREADS</span></tt></a> and <a class="reference internal" href="#Py_END_ALLOW_THREADS" title="Py_END_ALLOW_THREADS"><tt class="xref c c-macro docutils literal"><span class="pre">Py_END_ALLOW_THREADS</span></tt></a> macros is
acceptable.</p>
<p>The return value is an opaque &#8220;handle&#8221; to the thread state when
<a class="reference internal" href="#PyGILState_Ensure" title="PyGILState_Ensure"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Ensure()</span></tt></a> was called, and must be passed to
<a class="reference internal" href="#PyGILState_Release" title="PyGILState_Release"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Release()</span></tt></a> to ensure Python is left in the same state. Even
though recursive calls are allowed, these handles <em>cannot</em> be shared - each
unique call to <a class="reference internal" href="#PyGILState_Ensure" title="PyGILState_Ensure"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Ensure()</span></tt></a> must save the handle for its call
to <a class="reference internal" href="#PyGILState_Release" title="PyGILState_Release"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Release()</span></tt></a>.</p>
<p>When the function returns, the current thread will hold the GIL and be able
to call arbitrary Python code.  Failure is a fatal error.</p>
</dd></dl>

<dl class="function">
<dt id="PyGILState_Release">
void <tt class="descname">PyGILState_Release</tt><big>(</big>PyGILState_STATE<big>)</big><a class="headerlink" href="#PyGILState_Release" title="Permalink to this definition">¶</a></dt>
<dd><p>Release any resources previously acquired.  After this call, Python&#8217;s state will
be the same as it was prior to the corresponding <a class="reference internal" href="#PyGILState_Ensure" title="PyGILState_Ensure"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Ensure()</span></tt></a> call
(but generally this state will be unknown to the caller, hence the use of the
GILState API).</p>
<p>Every call to <a class="reference internal" href="#PyGILState_Ensure" title="PyGILState_Ensure"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Ensure()</span></tt></a> must be matched by a call to
<a class="reference internal" href="#PyGILState_Release" title="PyGILState_Release"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Release()</span></tt></a> on the same thread.</p>
</dd></dl>

<dl class="function">
<dt id="PyGILState_GetThisThreadState">
<a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a> <tt class="descname">PyGILState_GetThisThreadState</tt><big>(</big><big>)</big><a class="headerlink" href="#PyGILState_GetThisThreadState" title="Permalink to this definition">¶</a></dt>
<dd><p>Get the current thread state for this thread.  May return <tt class="docutils literal"><span class="pre">NULL</span></tt> if no
GILState API has been used on the current thread.  Note that the main thread
always has such a thread-state, even if no auto-thread-state call has been
made on the main thread.  This is mainly a helper/diagnostic function.</p>
</dd></dl>

<p>The following macros are normally used without a trailing semicolon; look for
example usage in the Python source distribution.</p>
<dl class="macro">
<dt id="Py_BEGIN_ALLOW_THREADS">
<tt class="descname">Py_BEGIN_ALLOW_THREADS</tt><a class="headerlink" href="#Py_BEGIN_ALLOW_THREADS" title="Permalink to this definition">¶</a></dt>
<dd><p>This macro expands to <tt class="docutils literal"><span class="pre">{</span> <span class="pre">PyThreadState</span> <span class="pre">*_save;</span> <span class="pre">_save</span> <span class="pre">=</span> <span class="pre">PyEval_SaveThread();</span></tt>.
Note that it contains an opening brace; it must be matched with a following
<a class="reference internal" href="#Py_END_ALLOW_THREADS" title="Py_END_ALLOW_THREADS"><tt class="xref c c-macro docutils literal"><span class="pre">Py_END_ALLOW_THREADS</span></tt></a> macro.  See above for further discussion of this
macro.  It is a no-op when thread support is disabled at compile time.</p>
</dd></dl>

<dl class="macro">
<dt id="Py_END_ALLOW_THREADS">
<tt class="descname">Py_END_ALLOW_THREADS</tt><a class="headerlink" href="#Py_END_ALLOW_THREADS" title="Permalink to this definition">¶</a></dt>
<dd><p>This macro expands to <tt class="docutils literal"><span class="pre">PyEval_RestoreThread(_save);</span> <span class="pre">}</span></tt>. Note that it contains
a closing brace; it must be matched with an earlier
<a class="reference internal" href="#Py_BEGIN_ALLOW_THREADS" title="Py_BEGIN_ALLOW_THREADS"><tt class="xref c c-macro docutils literal"><span class="pre">Py_BEGIN_ALLOW_THREADS</span></tt></a> macro.  See above for further discussion of
this macro.  It is a no-op when thread support is disabled at compile time.</p>
</dd></dl>

<dl class="macro">
<dt id="Py_BLOCK_THREADS">
<tt class="descname">Py_BLOCK_THREADS</tt><a class="headerlink" href="#Py_BLOCK_THREADS" title="Permalink to this definition">¶</a></dt>
<dd><p>This macro expands to <tt class="docutils literal"><span class="pre">PyEval_RestoreThread(_save);</span></tt>: it is equivalent to
<a class="reference internal" href="#Py_END_ALLOW_THREADS" title="Py_END_ALLOW_THREADS"><tt class="xref c c-macro docutils literal"><span class="pre">Py_END_ALLOW_THREADS</span></tt></a> without the closing brace.  It is a no-op when
thread support is disabled at compile time.</p>
</dd></dl>

<dl class="macro">
<dt id="Py_UNBLOCK_THREADS">
<tt class="descname">Py_UNBLOCK_THREADS</tt><a class="headerlink" href="#Py_UNBLOCK_THREADS" title="Permalink to this definition">¶</a></dt>
<dd><p>This macro expands to <tt class="docutils literal"><span class="pre">_save</span> <span class="pre">=</span> <span class="pre">PyEval_SaveThread();</span></tt>: it is equivalent to
<a class="reference internal" href="#Py_BEGIN_ALLOW_THREADS" title="Py_BEGIN_ALLOW_THREADS"><tt class="xref c c-macro docutils literal"><span class="pre">Py_BEGIN_ALLOW_THREADS</span></tt></a> without the opening brace and variable
declaration.  It is a no-op when thread support is disabled at compile time.</p>
</dd></dl>

</div>
<div class="section" id="low-level-api">
<h3>Low-level API<a class="headerlink" href="#low-level-api" title="Permalink to this headline">¶</a></h3>
<p>All of the following functions are only available when thread support is enabled
at compile time, and must be called only when the global interpreter lock has
been created.</p>
<dl class="function">
<dt id="PyInterpreterState_New">
<a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState">PyInterpreterState</a>* <tt class="descname">PyInterpreterState_New</tt><big>(</big><big>)</big><a class="headerlink" href="#PyInterpreterState_New" title="Permalink to this definition">¶</a></dt>
<dd><p>Create a new interpreter state object.  The global interpreter lock need not
be held, but may be held if it is necessary to serialize calls to this
function.</p>
</dd></dl>

<dl class="function">
<dt id="PyInterpreterState_Clear">
void <tt class="descname">PyInterpreterState_Clear</tt><big>(</big><a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState">PyInterpreterState</a><em>&nbsp;*interp</em><big>)</big><a class="headerlink" href="#PyInterpreterState_Clear" title="Permalink to this definition">¶</a></dt>
<dd><p>Reset all information in an interpreter state object.  The global interpreter
lock must be held.</p>
</dd></dl>

<dl class="function">
<dt id="PyInterpreterState_Delete">
void <tt class="descname">PyInterpreterState_Delete</tt><big>(</big><a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState">PyInterpreterState</a><em>&nbsp;*interp</em><big>)</big><a class="headerlink" href="#PyInterpreterState_Delete" title="Permalink to this definition">¶</a></dt>
<dd><p>Destroy an interpreter state object.  The global interpreter lock need not be
held.  The interpreter state must have been reset with a previous call to
<a class="reference internal" href="#PyInterpreterState_Clear" title="PyInterpreterState_Clear"><tt class="xref c c-func docutils literal"><span class="pre">PyInterpreterState_Clear()</span></tt></a>.</p>
</dd></dl>

<dl class="function">
<dt id="PyThreadState_New">
<a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a>* <tt class="descname">PyThreadState_New</tt><big>(</big><a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState">PyInterpreterState</a><em>&nbsp;*interp</em><big>)</big><a class="headerlink" href="#PyThreadState_New" title="Permalink to this definition">¶</a></dt>
<dd><p>Create a new thread state object belonging to the given interpreter object.
The global interpreter lock need not be held, but may be held if it is
necessary to serialize calls to this function.</p>
</dd></dl>

<dl class="function">
<dt id="PyThreadState_Clear">
void <tt class="descname">PyThreadState_Clear</tt><big>(</big><a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a><em>&nbsp;*tstate</em><big>)</big><a class="headerlink" href="#PyThreadState_Clear" title="Permalink to this definition">¶</a></dt>
<dd><p>Reset all information in a thread state object.  The global interpreter lock
must be held.</p>
</dd></dl>

<dl class="function">
<dt id="PyThreadState_Delete">
void <tt class="descname">PyThreadState_Delete</tt><big>(</big><a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a><em>&nbsp;*tstate</em><big>)</big><a class="headerlink" href="#PyThreadState_Delete" title="Permalink to this definition">¶</a></dt>
<dd><p>Destroy a thread state object.  The global interpreter lock need not be held.
The thread state must have been reset with a previous call to
<a class="reference internal" href="#PyThreadState_Clear" title="PyThreadState_Clear"><tt class="xref c c-func docutils literal"><span class="pre">PyThreadState_Clear()</span></tt></a>.</p>
</dd></dl>

<dl class="function">
<dt id="PyThreadState_GetDict">
<a class="reference internal" href="structures.html#PyObject" title="PyObject">PyObject</a>* <tt class="descname">PyThreadState_GetDict</tt><big>(</big><big>)</big><a class="headerlink" href="#PyThreadState_GetDict" title="Permalink to this definition">¶</a></dt>
<dd><em class="refcount">Return value: Borrowed reference.</em><p>Return a dictionary in which extensions can store thread-specific state
information.  Each extension should use a unique key to use to store state in
the dictionary.  It is okay to call this function when no current thread state
is available. If this function returns <em>NULL</em>, no exception has been raised and
the caller should assume no current thread state is available.</p>
</dd></dl>

<dl class="function">
<dt id="PyThreadState_SetAsyncExc">
int <tt class="descname">PyThreadState_SetAsyncExc</tt><big>(</big>long<em>&nbsp;id</em>, <a class="reference internal" href="structures.html#PyObject" title="PyObject">PyObject</a><em>&nbsp;*exc</em><big>)</big><a class="headerlink" href="#PyThreadState_SetAsyncExc" title="Permalink to this definition">¶</a></dt>
<dd><p>Asynchronously raise an exception in a thread. The <em>id</em> argument is the thread
id of the target thread; <em>exc</em> is the exception object to be raised. This
function does not steal any references to <em>exc</em>. To prevent naive misuse, you
must write your own C extension to call this.  Must be called with the GIL held.
Returns the number of thread states modified; this is normally one, but will be
zero if the thread id isn&#8217;t found.  If <em>exc</em> is <tt class="xref py py-const docutils literal"><span class="pre">NULL</span></tt>, the pending
exception (if any) for the thread is cleared. This raises no exceptions.</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_AcquireThread">
void <tt class="descname">PyEval_AcquireThread</tt><big>(</big><a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a><em>&nbsp;*tstate</em><big>)</big><a class="headerlink" href="#PyEval_AcquireThread" title="Permalink to this definition">¶</a></dt>
<dd><p>Acquire the global interpreter lock and set the current thread state to
<em>tstate</em>, which should not be <em>NULL</em>.  The lock must have been created earlier.
If this thread already has the lock, deadlock ensues.</p>
<p><a class="reference internal" href="#PyEval_RestoreThread" title="PyEval_RestoreThread"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_RestoreThread()</span></tt></a> is a higher-level function which is always
available (even when thread support isn&#8217;t enabled or when threads have
not been initialized).</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_ReleaseThread">
void <tt class="descname">PyEval_ReleaseThread</tt><big>(</big><a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a><em>&nbsp;*tstate</em><big>)</big><a class="headerlink" href="#PyEval_ReleaseThread" title="Permalink to this definition">¶</a></dt>
<dd><p>Reset the current thread state to <em>NULL</em> and release the global interpreter
lock.  The lock must have been created earlier and must be held by the current
thread.  The <em>tstate</em> argument, which must not be <em>NULL</em>, is only used to check
that it represents the current thread state &#8212; if it isn&#8217;t, a fatal error is
reported.</p>
<p><a class="reference internal" href="#PyEval_SaveThread" title="PyEval_SaveThread"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_SaveThread()</span></tt></a> is a higher-level function which is always
available (even when thread support isn&#8217;t enabled or when threads have
not been initialized).</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_AcquireLock">
void <tt class="descname">PyEval_AcquireLock</tt><big>(</big><big>)</big><a class="headerlink" href="#PyEval_AcquireLock" title="Permalink to this definition">¶</a></dt>
<dd><p>Acquire the global interpreter lock.  The lock must have been created earlier.
If this thread already has the lock, a deadlock ensues.</p>
<p class="deprecated">
<span class="versionmodified">Deprecated since version 3.2: </span>This function does not update the current thread state.  Please use
<a class="reference internal" href="#PyEval_RestoreThread" title="PyEval_RestoreThread"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_RestoreThread()</span></tt></a> or <a class="reference internal" href="#PyEval_AcquireThread" title="PyEval_AcquireThread"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_AcquireThread()</span></tt></a>
instead.</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_ReleaseLock">
void <tt class="descname">PyEval_ReleaseLock</tt><big>(</big><big>)</big><a class="headerlink" href="#PyEval_ReleaseLock" title="Permalink to this definition">¶</a></dt>
<dd><p>Release the global interpreter lock.  The lock must have been created earlier.</p>
<p class="deprecated">
<span class="versionmodified">Deprecated since version 3.2: </span>This function does not update the current thread state.  Please use
<a class="reference internal" href="#PyEval_SaveThread" title="PyEval_SaveThread"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_SaveThread()</span></tt></a> or <a class="reference internal" href="#PyEval_ReleaseThread" title="PyEval_ReleaseThread"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_ReleaseThread()</span></tt></a>
instead.</p>
</dd></dl>

</div>
</div>
<div class="section" id="sub-interpreter-support">
<h2>Sub-interpreter support<a class="headerlink" href="#sub-interpreter-support" title="Permalink to this headline">¶</a></h2>
<p>While in most uses, you will only embed a single Python interpreter, there
are cases where you need to create several independent interpreters in the
same process and perhaps even in the same thread.  Sub-interpreters allow
you to do that.  You can switch between sub-interpreters using the
<a class="reference internal" href="#PyThreadState_Swap" title="PyThreadState_Swap"><tt class="xref c c-func docutils literal"><span class="pre">PyThreadState_Swap()</span></tt></a> function.  You can create and destroy them
using the following functions:</p>
<dl class="function">
<dt id="Py_NewInterpreter">
<a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a>* <tt class="descname">Py_NewInterpreter</tt><big>(</big><big>)</big><a class="headerlink" href="#Py_NewInterpreter" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-21">Create a new sub-interpreter.  This is an (almost) totally separate environment
for the execution of Python code.  In particular, the new interpreter has
separate, independent versions of all imported modules, including the
fundamental modules <a class="reference internal" href="../library/builtins.html#module-builtins" title="builtins: The module that provides the built-in namespace."><tt class="xref py py-mod docutils literal"><span class="pre">builtins</span></tt></a>, <a class="reference internal" href="../library/__main__.html#module-__main__" title="__main__: The environment where the top-level script is run."><tt class="xref py py-mod docutils literal"><span class="pre">__main__</span></tt></a> and <a class="reference internal" href="../library/sys.html#module-sys" title="sys: Access system-specific parameters and functions."><tt class="xref py py-mod docutils literal"><span class="pre">sys</span></tt></a>.  The
table of loaded modules (<tt class="docutils literal"><span class="pre">sys.modules</span></tt>) and the module search path
(<tt class="docutils literal"><span class="pre">sys.path</span></tt>) are also separate.  The new environment has no <tt class="docutils literal"><span class="pre">sys.argv</span></tt>
variable.  It has new standard I/O stream file objects <tt class="docutils literal"><span class="pre">sys.stdin</span></tt>,
<tt class="docutils literal"><span class="pre">sys.stdout</span></tt> and <tt class="docutils literal"><span class="pre">sys.stderr</span></tt> (however these refer to the same underlying
file descriptors).</p>
<p>The return value points to the first thread state created in the new
sub-interpreter.  This thread state is made in the current thread state.
Note that no actual thread is created; see the discussion of thread states
below.  If creation of the new interpreter is unsuccessful, <em>NULL</em> is
returned; no exception is set since the exception state is stored in the
current thread state and there may not be a current thread state.  (Like all
other Python/C API functions, the global interpreter lock must be held before
calling this function and is still held when it returns; however, unlike most
other Python/C API functions, there needn&#8217;t be a current thread state on
entry.)</p>
<p id="index-22">Extension modules are shared between (sub-)interpreters as follows: the first
time a particular extension is imported, it is initialized normally, and a
(shallow) copy of its module&#8217;s dictionary is squirreled away.  When the same
extension is imported by another (sub-)interpreter, a new module is initialized
and filled with the contents of this copy; the extension&#8217;s <tt class="docutils literal"><span class="pre">init</span></tt> function is
not called.  Note that this is different from what happens when an extension is
imported after the interpreter has been completely re-initialized by calling
<a class="reference internal" href="#Py_Finalize" title="Py_Finalize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Finalize()</span></tt></a> and <a class="reference internal" href="#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Initialize()</span></tt></a>; in that case, the extension&#8217;s
<tt class="docutils literal"><span class="pre">initmodule</span></tt> function <em>is</em> called again.</p>
<span class="target" id="index-23"></span></dd></dl>

<dl class="function">
<dt id="Py_EndInterpreter">
void <tt class="descname">Py_EndInterpreter</tt><big>(</big><a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a><em>&nbsp;*tstate</em><big>)</big><a class="headerlink" href="#Py_EndInterpreter" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-24">Destroy the (sub-)interpreter represented by the given thread state. The given
thread state must be the current thread state.  See the discussion of thread
states below.  When the call returns, the current thread state is <em>NULL</em>.  All
thread states associated with this interpreter are destroyed.  (The global
interpreter lock must be held before calling this function and is still held
when it returns.)  <a class="reference internal" href="#Py_Finalize" title="Py_Finalize"><tt class="xref c c-func docutils literal"><span class="pre">Py_Finalize()</span></tt></a> will destroy all sub-interpreters that
haven&#8217;t been explicitly destroyed at that point.</p>
</dd></dl>

<div class="section" id="bugs-and-caveats">
<h3>Bugs and caveats<a class="headerlink" href="#bugs-and-caveats" title="Permalink to this headline">¶</a></h3>
<p>Because sub-interpreters (and the main interpreter) are part of the same
process, the insulation between them isn&#8217;t perfect &#8212; for example, using
low-level file operations like  <a class="reference internal" href="../library/os.html#os.close" title="os.close"><tt class="xref py py-func docutils literal"><span class="pre">os.close()</span></tt></a> they can
(accidentally or maliciously) affect each other&#8217;s open files.  Because of the
way extensions are shared between (sub-)interpreters, some extensions may not
work properly; this is especially likely when the extension makes use of
(static) global variables, or when the extension manipulates its module&#8217;s
dictionary after its initialization.  It is possible to insert objects created
in one sub-interpreter into a namespace of another sub-interpreter; this should
be done with great care to avoid sharing user-defined functions, methods,
instances or classes between sub-interpreters, since import operations executed
by such objects may affect the wrong (sub-)interpreter&#8217;s dictionary of loaded
modules.</p>
<p>Also note that combining this functionality with <tt class="xref c c-func docutils literal"><span class="pre">PyGILState_*()</span></tt> APIs
is delicate, because these APIs assume a bijection between Python thread states
and OS-level threads, an assumption broken by the presence of sub-interpreters.
It is highly recommended that you don&#8217;t switch sub-interpreters between a pair
of matching <a class="reference internal" href="#PyGILState_Ensure" title="PyGILState_Ensure"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Ensure()</span></tt></a> and <a class="reference internal" href="#PyGILState_Release" title="PyGILState_Release"><tt class="xref c c-func docutils literal"><span class="pre">PyGILState_Release()</span></tt></a> calls.
Furthermore, extensions (such as <a class="reference internal" href="../library/ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><tt class="xref py py-mod docutils literal"><span class="pre">ctypes</span></tt></a>) using these APIs to allow calling
of Python code from non-Python created threads will probably be broken when using
sub-interpreters.</p>
</div>
</div>
<div class="section" id="asynchronous-notifications">
<h2>Asynchronous Notifications<a class="headerlink" href="#asynchronous-notifications" title="Permalink to this headline">¶</a></h2>
<p>A mechanism is provided to make asynchronous notifications to the main
interpreter thread.  These notifications take the form of a function
pointer and a void argument.</p>
<p id="index-25">Every check interval, when the global interpreter lock is released and
reacquired, Python will also call any such provided functions.  This can be used
for example by asynchronous IO handlers.  The notification can be scheduled from
a worker thread and the actual call than made at the earliest convenience by the
main thread where it has possession of the global interpreter lock and can
perform any Python API calls.</p>
<dl class="function">
<dt id="Py_AddPendingCall">
int <tt class="descname">Py_AddPendingCall</tt><big>(</big>int (*func)(void<em>&nbsp;*)</em>, void<em>&nbsp;*arg</em><big>)</big><a class="headerlink" href="#Py_AddPendingCall" title="Permalink to this definition">¶</a></dt>
<dd><p id="index-26">Post a notification to the Python main thread.  If successful, <em>func</em> will be
called with the argument <em>arg</em> at the earliest convenience.  <em>func</em> will be
called having the global interpreter lock held and can thus use the full
Python API and can take any action such as setting object attributes to
signal IO completion.  It must return 0 on success, or -1 signalling an
exception.  The notification function won&#8217;t be interrupted to perform another
asynchronous notification recursively, but it can still be interrupted to
switch threads if the global interpreter lock is released, for example, if it
calls back into Python code.</p>
<p>This function returns 0 on success in which case the notification has been
scheduled.  Otherwise, for example if the notification buffer is full, it
returns -1 without setting any exception.</p>
<p>This function can be called on any thread, be it a Python thread or some
other system thread.  If it is a Python thread, it doesn&#8217;t matter if it holds
the global interpreter lock or not.</p>
<p class="versionadded">
<span class="versionmodified">New in version 3.1.</span></p>
</dd></dl>

</div>
<div class="section" id="profiling-and-tracing">
<span id="profiling"></span><h2>Profiling and Tracing<a class="headerlink" href="#profiling-and-tracing" title="Permalink to this headline">¶</a></h2>
<p>The Python interpreter provides some low-level support for attaching profiling
and execution tracing facilities.  These are used for profiling, debugging, and
coverage analysis tools.</p>
<p>This C interface allows the profiling or tracing code to avoid the overhead of
calling through Python-level callable objects, making a direct C function call
instead.  The essential attributes of the facility have not changed; the
interface allows trace functions to be installed per-thread, and the basic
events reported to the trace function are the same as had been reported to the
Python-level trace functions in previous versions.</p>
<dl class="type">
<dt id="Py_tracefunc">
int <tt class="descname">(*Py_tracefunc)</tt><big>(</big><a class="reference internal" href="structures.html#PyObject" title="PyObject">PyObject</a><em>&nbsp;*obj</em>, PyFrameObject<em>&nbsp;*frame</em>, int<em>&nbsp;what</em>, <a class="reference internal" href="structures.html#PyObject" title="PyObject">PyObject</a><em>&nbsp;*arg</em><big>)</big><a class="headerlink" href="#Py_tracefunc" title="Permalink to this definition">¶</a></dt>
<dd><p>The type of the trace function registered using <a class="reference internal" href="#PyEval_SetProfile" title="PyEval_SetProfile"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_SetProfile()</span></tt></a> and
<a class="reference internal" href="#PyEval_SetTrace" title="PyEval_SetTrace"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_SetTrace()</span></tt></a>. The first parameter is the object passed to the
registration function as <em>obj</em>, <em>frame</em> is the frame object to which the event
pertains, <em>what</em> is one of the constants <tt class="xref py py-const docutils literal"><span class="pre">PyTrace_CALL</span></tt>,
<tt class="xref py py-const docutils literal"><span class="pre">PyTrace_EXCEPTION</span></tt>, <tt class="xref py py-const docutils literal"><span class="pre">PyTrace_LINE</span></tt>, <tt class="xref py py-const docutils literal"><span class="pre">PyTrace_RETURN</span></tt>,
<tt class="xref py py-const docutils literal"><span class="pre">PyTrace_C_CALL</span></tt>, <tt class="xref py py-const docutils literal"><span class="pre">PyTrace_C_EXCEPTION</span></tt>, or
<tt class="xref py py-const docutils literal"><span class="pre">PyTrace_C_RETURN</span></tt>, and <em>arg</em> depends on the value of <em>what</em>:</p>
<table border="1" class="docutils">
<colgroup>
<col width="44%" />
<col width="56%" />
</colgroup>
<thead valign="bottom">
<tr><th class="head">Value of <em>what</em></th>
<th class="head">Meaning of <em>arg</em></th>
</tr>
</thead>
<tbody valign="top">
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PyTrace_CALL</span></tt></td>
<td>Always <em>NULL</em>.</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PyTrace_EXCEPTION</span></tt></td>
<td>Exception information as returned by
<a class="reference internal" href="../library/sys.html#sys.exc_info" title="sys.exc_info"><tt class="xref py py-func docutils literal"><span class="pre">sys.exc_info()</span></tt></a>.</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PyTrace_LINE</span></tt></td>
<td>Always <em>NULL</em>.</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PyTrace_RETURN</span></tt></td>
<td>Value being returned to the caller,
or <em>NULL</em> if caused by an exception.</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PyTrace_C_CALL</span></tt></td>
<td>Function object being called.</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PyTrace_C_EXCEPTION</span></tt></td>
<td>Function object being called.</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PyTrace_C_RETURN</span></tt></td>
<td>Function object being called.</td>
</tr>
</tbody>
</table>
</dd></dl>

<dl class="var">
<dt id="PyTrace_CALL">
int <tt class="descname">PyTrace_CALL</tt><a class="headerlink" href="#PyTrace_CALL" title="Permalink to this definition">¶</a></dt>
<dd><p>The value of the <em>what</em> parameter to a <a class="reference internal" href="#Py_tracefunc" title="Py_tracefunc"><tt class="xref c c-type docutils literal"><span class="pre">Py_tracefunc</span></tt></a> function when a new
call to a function or method is being reported, or a new entry into a generator.
Note that the creation of the iterator for a generator function is not reported
as there is no control transfer to the Python bytecode in the corresponding
frame.</p>
</dd></dl>

<dl class="var">
<dt id="PyTrace_EXCEPTION">
int <tt class="descname">PyTrace_EXCEPTION</tt><a class="headerlink" href="#PyTrace_EXCEPTION" title="Permalink to this definition">¶</a></dt>
<dd><p>The value of the <em>what</em> parameter to a <a class="reference internal" href="#Py_tracefunc" title="Py_tracefunc"><tt class="xref c c-type docutils literal"><span class="pre">Py_tracefunc</span></tt></a> function when an
exception has been raised.  The callback function is called with this value for
<em>what</em> when after any bytecode is processed after which the exception becomes
set within the frame being executed.  The effect of this is that as exception
propagation causes the Python stack to unwind, the callback is called upon
return to each frame as the exception propagates.  Only trace functions receives
these events; they are not needed by the profiler.</p>
</dd></dl>

<dl class="var">
<dt id="PyTrace_LINE">
int <tt class="descname">PyTrace_LINE</tt><a class="headerlink" href="#PyTrace_LINE" title="Permalink to this definition">¶</a></dt>
<dd><p>The value passed as the <em>what</em> parameter to a trace function (but not a
profiling function) when a line-number event is being reported.</p>
</dd></dl>

<dl class="var">
<dt id="PyTrace_RETURN">
int <tt class="descname">PyTrace_RETURN</tt><a class="headerlink" href="#PyTrace_RETURN" title="Permalink to this definition">¶</a></dt>
<dd><p>The value for the <em>what</em> parameter to <a class="reference internal" href="#Py_tracefunc" title="Py_tracefunc"><tt class="xref c c-type docutils literal"><span class="pre">Py_tracefunc</span></tt></a> functions when a
call is returning without propagating an exception.</p>
</dd></dl>

<dl class="var">
<dt id="PyTrace_C_CALL">
int <tt class="descname">PyTrace_C_CALL</tt><a class="headerlink" href="#PyTrace_C_CALL" title="Permalink to this definition">¶</a></dt>
<dd><p>The value for the <em>what</em> parameter to <a class="reference internal" href="#Py_tracefunc" title="Py_tracefunc"><tt class="xref c c-type docutils literal"><span class="pre">Py_tracefunc</span></tt></a> functions when a C
function is about to be called.</p>
</dd></dl>

<dl class="var">
<dt id="PyTrace_C_EXCEPTION">
int <tt class="descname">PyTrace_C_EXCEPTION</tt><a class="headerlink" href="#PyTrace_C_EXCEPTION" title="Permalink to this definition">¶</a></dt>
<dd><p>The value for the <em>what</em> parameter to <a class="reference internal" href="#Py_tracefunc" title="Py_tracefunc"><tt class="xref c c-type docutils literal"><span class="pre">Py_tracefunc</span></tt></a> functions when a C
function has raised an exception.</p>
</dd></dl>

<dl class="var">
<dt id="PyTrace_C_RETURN">
int <tt class="descname">PyTrace_C_RETURN</tt><a class="headerlink" href="#PyTrace_C_RETURN" title="Permalink to this definition">¶</a></dt>
<dd><p>The value for the <em>what</em> parameter to <a class="reference internal" href="#Py_tracefunc" title="Py_tracefunc"><tt class="xref c c-type docutils literal"><span class="pre">Py_tracefunc</span></tt></a> functions when a C
function has returned.</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_SetProfile">
void <tt class="descname">PyEval_SetProfile</tt><big>(</big><a class="reference internal" href="#Py_tracefunc" title="Py_tracefunc">Py_tracefunc</a><em>&nbsp;func</em>, <a class="reference internal" href="structures.html#PyObject" title="PyObject">PyObject</a><em>&nbsp;*obj</em><big>)</big><a class="headerlink" href="#PyEval_SetProfile" title="Permalink to this definition">¶</a></dt>
<dd><p>Set the profiler function to <em>func</em>.  The <em>obj</em> parameter is passed to the
function as its first parameter, and may be any Python object, or <em>NULL</em>.  If
the profile function needs to maintain state, using a different value for <em>obj</em>
for each thread provides a convenient and thread-safe place to store it.  The
profile function is called for all monitored events except the line-number
events.</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_SetTrace">
void <tt class="descname">PyEval_SetTrace</tt><big>(</big><a class="reference internal" href="#Py_tracefunc" title="Py_tracefunc">Py_tracefunc</a><em>&nbsp;func</em>, <a class="reference internal" href="structures.html#PyObject" title="PyObject">PyObject</a><em>&nbsp;*obj</em><big>)</big><a class="headerlink" href="#PyEval_SetTrace" title="Permalink to this definition">¶</a></dt>
<dd><p>Set the tracing function to <em>func</em>.  This is similar to
<a class="reference internal" href="#PyEval_SetProfile" title="PyEval_SetProfile"><tt class="xref c c-func docutils literal"><span class="pre">PyEval_SetProfile()</span></tt></a>, except the tracing function does receive line-number
events.</p>
</dd></dl>

<dl class="function">
<dt id="PyEval_GetCallStats">
<a class="reference internal" href="structures.html#PyObject" title="PyObject">PyObject</a>* <tt class="descname">PyEval_GetCallStats</tt><big>(</big><a class="reference internal" href="structures.html#PyObject" title="PyObject">PyObject</a><em>&nbsp;*self</em><big>)</big><a class="headerlink" href="#PyEval_GetCallStats" title="Permalink to this definition">¶</a></dt>
<dd><p>Return a tuple of function call counts.  There are constants defined for the
positions within the tuple:</p>
<table border="1" class="docutils">
<colgroup>
<col width="82%" />
<col width="18%" />
</colgroup>
<thead valign="bottom">
<tr><th class="head">Name</th>
<th class="head">Value</th>
</tr>
</thead>
<tbody valign="top">
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_ALL</span></tt></td>
<td>0</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_FUNCTION</span></tt></td>
<td>1</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_FAST_FUNCTION</span></tt></td>
<td>2</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_FASTER_FUNCTION</span></tt></td>
<td>3</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_METHOD</span></tt></td>
<td>4</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_BOUND_METHOD</span></tt></td>
<td>5</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_CFUNCTION</span></tt></td>
<td>6</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_TYPE</span></tt></td>
<td>7</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_GENERATOR</span></tt></td>
<td>8</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_OTHER</span></tt></td>
<td>9</td>
</tr>
<tr><td><tt class="xref py py-const docutils literal"><span class="pre">PCALL_POP</span></tt></td>
<td>10</td>
</tr>
</tbody>
</table>
<p><tt class="xref py py-const docutils literal"><span class="pre">PCALL_FAST_FUNCTION</span></tt> means no argument tuple needs to be created.
<tt class="xref py py-const docutils literal"><span class="pre">PCALL_FASTER_FUNCTION</span></tt> means that the fast-path frame setup code is used.</p>
<p>If there is a method call where the call can be optimized by changing
the argument tuple and calling the function directly, it gets recorded
twice.</p>
<p>This function is only present if Python is compiled with <tt class="xref py py-const docutils literal"><span class="pre">CALL_PROFILE</span></tt>
defined.</p>
</dd></dl>

</div>
<div class="section" id="advanced-debugger-support">
<span id="advanced-debugging"></span><h2>Advanced Debugger Support<a class="headerlink" href="#advanced-debugger-support" title="Permalink to this headline">¶</a></h2>
<p>These functions are only intended to be used by advanced debugging tools.</p>
<dl class="function">
<dt id="PyInterpreterState_Head">
<a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState">PyInterpreterState</a>* <tt class="descname">PyInterpreterState_Head</tt><big>(</big><big>)</big><a class="headerlink" href="#PyInterpreterState_Head" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the interpreter state object at the head of the list of all such objects.</p>
</dd></dl>

<dl class="function">
<dt id="PyInterpreterState_Next">
<a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState">PyInterpreterState</a>* <tt class="descname">PyInterpreterState_Next</tt><big>(</big><a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState">PyInterpreterState</a><em>&nbsp;*interp</em><big>)</big><a class="headerlink" href="#PyInterpreterState_Next" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the next interpreter state object after <em>interp</em> from the list of all
such objects.</p>
</dd></dl>

<dl class="function">
<dt id="PyInterpreterState_ThreadHead">
<a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a> * <tt class="descname">PyInterpreterState_ThreadHead</tt><big>(</big><a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState">PyInterpreterState</a><em>&nbsp;*interp</em><big>)</big><a class="headerlink" href="#PyInterpreterState_ThreadHead" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the a pointer to the first <a class="reference internal" href="#PyThreadState" title="PyThreadState"><tt class="xref c c-type docutils literal"><span class="pre">PyThreadState</span></tt></a> object in the list of
threads associated with the interpreter <em>interp</em>.</p>
</dd></dl>

<dl class="function">
<dt id="PyThreadState_Next">
<a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a>* <tt class="descname">PyThreadState_Next</tt><big>(</big><a class="reference internal" href="#PyThreadState" title="PyThreadState">PyThreadState</a><em>&nbsp;*tstate</em><big>)</big><a class="headerlink" href="#PyThreadState_Next" title="Permalink to this definition">¶</a></dt>
<dd><p>Return the next thread state object after <em>tstate</em> from the list of all such
objects belonging to the same <a class="reference internal" href="#PyInterpreterState" title="PyInterpreterState"><tt class="xref c c-type docutils literal"><span class="pre">PyInterpreterState</span></tt></a> object.</p>
</dd></dl>

</div>
</div>


          </div>
        </div>
      </div>
      <div class="sphinxsidebar">
        <div class="sphinxsidebarwrapper">
  <h3><a href="../contents.html">Table Of Contents</a></h3>
  <ul>
<li><a class="reference internal" href="#">Initialization, Finalization, and Threads</a><ul>
<li><a class="reference internal" href="#initializing-and-finalizing-the-interpreter">Initializing and finalizing the interpreter</a></li>
<li><a class="reference internal" href="#process-wide-parameters">Process-wide parameters</a></li>
<li><a class="reference internal" href="#thread-state-and-the-global-interpreter-lock">Thread State and the Global Interpreter Lock</a><ul>
<li><a class="reference internal" href="#releasing-the-gil-from-extension-code">Releasing the GIL from extension code</a></li>
<li><a class="reference internal" href="#non-python-created-threads">Non-Python created threads</a></li>
<li><a class="reference internal" href="#high-level-api">High-level API</a></li>
<li><a class="reference internal" href="#low-level-api">Low-level API</a></li>
</ul>
</li>
<li><a class="reference internal" href="#sub-interpreter-support">Sub-interpreter support</a><ul>
<li><a class="reference internal" href="#bugs-and-caveats">Bugs and caveats</a></li>
</ul>
</li>
<li><a class="reference internal" href="#asynchronous-notifications">Asynchronous Notifications</a></li>
<li><a class="reference internal" href="#profiling-and-tracing">Profiling and Tracing</a></li>
<li><a class="reference internal" href="#advanced-debugger-support">Advanced Debugger Support</a></li>
</ul>
</li>
</ul>

  <h4>Previous topic</h4>
  <p class="topless"><a href="code.html"
                        title="previous chapter">Code Objects</a></p>
  <h4>Next topic</h4>
  <p class="topless"><a href="memory.html"
                        title="next chapter">Memory Management</a></p>
<h3>This Page</h3>
<ul class="this-page-menu">
  <li><a href="../bugs.html">Report a Bug</a></li>
  <li><a href="../_sources/c-api/init.txt"
         rel="nofollow">Show Source</a></li>
</ul>

<div id="searchbox" style="display: none">
  <h3>Quick search</h3>
    <form class="search" action="../search.html" method="get">
      <input type="text" name="q" size="18" />
      <input type="submit" value="Go" />
      <input type="hidden" name="check_keywords" value="yes" />
      <input type="hidden" name="area" value="default" />
    </form>
    <p class="searchtip" style="font-size: 90%">
    Enter search terms or a module, class or function name.
    </p>
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
        </div>
      </div>
      <div class="clearer"></div>
    </div>
    <div class="related">
      <h3>Navigation</h3>
      <ul>
        <li class="right" style="margin-right: 10px">
          <a href="../genindex.html" title="General Index"
             >index</a></li>
        <li class="right" >
          <a href="../py-modindex.html" title="Python Module Index"
             >modules</a> |</li>
        <li class="right" >
          <a href="memory.html" title="Memory Management"
             >next</a> |</li>
        <li class="right" >
          <a href="code.html" title="Code Objects"
             >previous</a> |</li>
        <li><img src="../_static/py.png" alt=""
                 style="vertical-align: middle; margin-top: -1px"/></li>
        <li><a href="../index.html">Python v3.2.2 documentation</a> &raquo;</li>

          <li><a href="index.html" >Python/C API Reference Manual</a> &raquo;</li> 
      </ul>
    </div>
    <div class="footer">
    &copy; <a href="../copyright.html">Copyright</a> 1990-2011, Python Software Foundation.
    <br />
    The Python Software Foundation is a non-profit corporation.  
    <a href="http://www.python.org/psf/donations/">Please donate.</a>
    <br />
    Last updated on Sep 04, 2011.
    <a href="../bugs.html">Found a bug</a>?
    <br />
    Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.0.7.
    </div>

  </body>
</html>