sam-t2/devtools/amd64/html/python/api/threads.html

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<html>
<head>
<link rel="STYLESHEET" href="api.css" type='text/css' />
<link rel="SHORTCUT ICON" href="../icons/pyfav.png" type="image/png" />
<link rel='start' href='../index.html' title='Python Documentation Index' />
<link rel="first" href="api.html" title='Python/C API Reference Manual' />
<link rel='contents' href='contents.html' title="Contents" />
<link rel='index' href='genindex.html' title='Index' />
<link rel='last' href='about.html' title='About this document...' />
<link rel='help' href='about.html' title='About this document...' />
<link rel="next" href="profiling.html" />
<link rel="prev" href="initialization.html" />
<link rel="parent" href="initialization.html" />
<link rel="next" href="profiling.html" />
<meta name='aesop' content='information' />
<title>8.1 Thread State and the Global Interpreter Lock </title>
</head>
<body>
<DIV CLASS="navigation">
<div id='top-navigation-panel' xml:id='top-navigation-panel'>
<table align="center" width="100%" cellpadding="0" cellspacing="2">
<tr>
<td class='online-navigation'><a rel="prev" title="8. Initialization, Finalization, and"
  href="initialization.html"><img src='../icons/previous.png'
  border='0' height='32'  alt='Previous Page' width='32' /></A></td>
<td class='online-navigation'><a rel="parent" title="8. Initialization, Finalization, and"
  href="initialization.html"><img src='../icons/up.png'
  border='0' height='32'  alt='Up One Level' width='32' /></A></td>
<td class='online-navigation'><a rel="next" title="8.2 Profiling and Tracing"
  href="profiling.html"><img src='../icons/next.png'
  border='0' height='32'  alt='Next Page' width='32' /></A></td>
<td align="center" width="100%">Python/C API Reference Manual</td>
<td class='online-navigation'><a rel="contents" title="Table of Contents"
  href="contents.html"><img src='../icons/contents.png'
  border='0' height='32'  alt='Contents' width='32' /></A></td>
<td class='online-navigation'><img src='../icons/blank.png'
  border='0' height='32'  alt='' width='32' /></td>
<td class='online-navigation'><a rel="index" title="Index"
  href="genindex.html"><img src='../icons/index.png'
  border='0' height='32'  alt='Index' width='32' /></A></td>
</tr></table>
<div class='online-navigation'>
<b class="navlabel">Previous:</b>
<a class="sectref" rel="prev" href="initialization.html">8. Initialization, Finalization, and</A>
<b class="navlabel">Up:</b>
<a class="sectref" rel="parent" href="initialization.html">8. Initialization, Finalization, and</A>
<b class="navlabel">Next:</b>
<a class="sectref" rel="next" href="profiling.html">8.2 Profiling and Tracing</A>
</div>
<hr /></div>
</DIV>
<!--End of Navigation Panel-->

<H1><A NAME="SECTION0010100000000000000000"></A><A NAME="threads"></A>
<BR>
8.1 Thread State and the Global Interpreter Lock

</H1>

<P>
<a id='l2h-865' xml:id='l2h-865'></a>

<P>
The Python interpreter is not fully thread safe.  In order to support
multi-threaded Python programs, there's a global lock 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>
Therefore, the rule exists that only the thread that has acquired the
global interpreter lock may operate on Python objects or call Python/C
API functions.  In order to support multi-threaded Python programs,
the interpreter regularly releases and reacquires the lock -- by
default, every 100 bytecode instructions (this can be changed with
<a id='l2h-838' xml:id='l2h-838'></a><tt class="function">sys.setcheckinterval()</tt>).  The lock is also released and
reacquired around potentially blocking I/O operations like reading or
writing a file, so that other threads can run while the thread that
requests the I/O is waiting for the I/O operation to complete.

<P>
The Python interpreter needs to keep some bookkeeping information
separate per thread -- for this it uses a data structure called
<tt class="ctype">PyThreadState</tt><a id='l2h-866' xml:id='l2h-866'></a>.  There's one global
variable, however: the pointer to the current
<tt class="ctype">PyThreadState</tt><a id='l2h-867' xml:id='l2h-867'></a> structure.  While most
thread packages have a way to store ``per-thread global data,''
Python's internal platform independent thread abstraction doesn't
support this yet.  Therefore, the current thread state must be
manipulated explicitly.

<P>
This is easy enough in most cases.  Most code manipulating the global
interpreter lock has the following simple structure:

<P>
<div class="verbatim"><pre>
Save the thread state in a local variable.
Release the interpreter lock.
...Do some blocking I/O operation...
Reacquire the interpreter lock.
Restore the thread state from the local variable.
</pre></div>

<P>
This is so common that a pair of macros exists to simplify it:

<P>
<div class="verbatim"><pre>
Py_BEGIN_ALLOW_THREADS
...Do some blocking I/O operation...
Py_END_ALLOW_THREADS
</pre></div>

<P>
The
Py_BEGIN_ALLOW_THREADS<a id='l2h-868' xml:id='l2h-868'></a>
macro opens a new block and declares a hidden local variable; the
Py_END_ALLOW_THREADS<a id='l2h-869' xml:id='l2h-869'></a>
macro closes the block.  Another advantage of using these two macros
is that when Python is compiled without thread support, they are
defined empty, thus saving the thread state and lock manipulations.

<P>
When thread support is enabled, the block above expands to the
following code:

<P>
<div class="verbatim"><pre>
    PyThreadState *_save;

    _save = PyEval_SaveThread();
    ...Do some blocking I/O operation...
    PyEval_RestoreThread(_save);
</pre></div>

<P>
Using even lower level primitives, we can get roughly the same effect
as follows:

<P>
<div class="verbatim"><pre>
    PyThreadState *_save;

    _save = PyThreadState_Swap(NULL);
    PyEval_ReleaseLock();
    ...Do some blocking I/O operation...
    PyEval_AcquireLock();
    PyThreadState_Swap(_save);
</pre></div>

<P>
There are some subtle differences; in particular,
<tt class="cfunction">PyEval_RestoreThread()</tt><a id='l2h-870' xml:id='l2h-870'></a> saves
and restores the value of the  global variable
<tt class="cdata">errno</tt><a id='l2h-871' xml:id='l2h-871'></a>, since the lock manipulation does not
guarantee that <tt class="cdata">errno</tt> is left alone.  Also, when thread support
is disabled,
<tt class="cfunction">PyEval_SaveThread()</tt><a id='l2h-872' xml:id='l2h-872'></a> and
<tt class="cfunction">PyEval_RestoreThread()</tt> don't manipulate the lock; in this
case, <tt class="cfunction">PyEval_ReleaseLock()</tt><a id='l2h-873' xml:id='l2h-873'></a> and
<tt class="cfunction">PyEval_AcquireLock()</tt><a id='l2h-874' xml:id='l2h-874'></a> are not
available.  This is done so that dynamically loaded extensions
compiled with thread support enabled can be loaded by an interpreter
that was compiled with disabled thread support.

<P>
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>
Why am I going on with so much detail about this?  Because when
threads are created from C, they don't have the global interpreter
lock, nor is there a thread state data structure for them.  Such
threads must bootstrap themselves into existence, by first creating a
thread state data structure, then acquiring the lock, and finally
storing their thread state pointer, before they can start using the
Python/C API.  When they are done, they should reset the thread state
pointer, release the lock, and finally free their thread state data
structure.

<P>
Beginning with version 2.3, threads can now take advantage of the
<tt class="cfunction">PyGILState_*()</tt> functions to do all of the above
automatically.  The typical idiom for calling into Python from a C
thread is now:

<P>
<div class="verbatim"><pre>
    PyGILState_STATE gstate;
    gstate = PyGILState_Ensure();

    /* Perform Python actions here.  */
    result = CallSomeFunction();
    /* evaluate result */

    /* Release the thread. No Python API allowed beyond this point. */
    PyGILState_Release(gstate);
</pre></div>

<P>
Note that the <tt class="cfunction">PyGILState_*()</tt> functions assume there is only
one global  interpreter (created automatically by
<tt class="cfunction">Py_Initialize()</tt>).  Python still supports the creation of
additional interpreters  (using <tt class="cfunction">Py_NewInterpreter()</tt>), but
mixing multiple interpreters and the <tt class="cfunction">PyGILState_*()</tt> API is
unsupported.

<P>
<dl><dt><b><tt class="ctype"><a id='l2h-839' xml:id='l2h-839'>PyInterpreterState</a></tt></b></dt>
<dd>
  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>
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.
</dl>

<P>
<dl><dt><b><tt class="ctype"><a id='l2h-840' xml:id='l2h-840'>PyThreadState</a></tt></b></dt>
<dd>
  This data structure represents the state of a single thread.  The
  only public data member is <tt class="ctype">PyInterpreterState
  *</tt><tt class="member">interp</tt>, which points to this thread's interpreter state.
</dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-841' xml:id='l2h-841' class="cfunction">PyEval_InitThreads</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
  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="cfunction">PyEval_ReleaseLock()</tt><a id='l2h-875' xml:id='l2h-875'></a> or
  <code>PyEval_ReleaseThread(<var>tstate</var>)</code><a id='l2h-876' xml:id='l2h-876'></a>.
  It is not needed before calling
  <tt class="cfunction">PyEval_SaveThread()</tt><a id='l2h-877' xml:id='l2h-877'></a> or
  <tt class="cfunction">PyEval_RestoreThread()</tt><a id='l2h-878' xml:id='l2h-878'></a>.

<P>
This is a no-op when called for a second time.  It is safe to call
  this function before calling
  <tt class="cfunction">Py_Initialize()</tt><a id='l2h-879' xml:id='l2h-879'></a>.

<P>
When only the main thread exists, no lock 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 lock, it also acquires it.  Before the Python
  <tt class="module">thread</tt><a id='l2h-880' xml:id='l2h-880'></a> module creates a new thread,
  knowing that either it has the lock or the lock hasn't been created
  yet, it calls <tt class="cfunction">PyEval_InitThreads()</tt>.  When this call
  returns, it is guaranteed that the lock has been created and that the
  calling thread has acquired it.

<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>
This function is not available when thread support is disabled at
  compile time.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>int&nbsp;<b><tt id='l2h-842' xml:id='l2h-842' class="cfunction">PyEval_ThreadsInitialized</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
  Returns a non-zero value if <tt class="cfunction">PyEval_InitThreads()</tt> has been
  called.  This function can be called without holding the lock, 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.
<span class="versionnote">New in version 2.4.</span>

</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-843' xml:id='l2h-843' class="cfunction">PyEval_AcquireLock</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
  Acquire the global interpreter lock.  The lock must have been
  created earlier.  If this thread already has the lock, a deadlock
  ensues.  This function is not available when thread support is
  disabled at compile time.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-844' xml:id='l2h-844' class="cfunction">PyEval_ReleaseLock</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
  Release the global interpreter lock.  The lock must have been
  created earlier.  This function is not available when thread support
  is disabled at compile time.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-845' xml:id='l2h-845' class="cfunction">PyEval_AcquireThread</tt></b>(</nobr></td><td>PyThreadState *<var>tstate</var>)</td></tr></table></dt>
<dd>
  Acquire the global interpreter lock and set the current thread
  state to <var>tstate</var>, which should not be <tt class="constant">NULL</tt>.  The lock must
  have been created earlier.  If this thread already has the lock,
  deadlock ensues.  This function is not available when thread support
  is disabled at compile time.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-846' xml:id='l2h-846' class="cfunction">PyEval_ReleaseThread</tt></b>(</nobr></td><td>PyThreadState *<var>tstate</var>)</td></tr></table></dt>
<dd>
  Reset the current thread state to <tt class="constant">NULL</tt> and release the global
  interpreter lock.  The lock must have been created earlier and must
  be held by the current thread.  The <var>tstate</var> argument, which
  must not be <tt class="constant">NULL</tt>, is only used to check that it represents the
  current thread state -- if it isn't, a fatal error is reported.
  This function is not available when thread support is disabled at
  compile time.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>PyThreadState*&nbsp;<b><tt id='l2h-847' xml:id='l2h-847' class="cfunction">PyEval_SaveThread</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
  Release the interpreter lock (if it has been created and thread
  support is enabled) and reset the thread state to <tt class="constant">NULL</tt>, returning
  the previous thread state (which is not <tt class="constant">NULL</tt>).  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.)
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-848' xml:id='l2h-848' class="cfunction">PyEval_RestoreThread</tt></b>(</nobr></td><td>PyThreadState *<var>tstate</var>)</td></tr></table></dt>
<dd>
  Acquire the interpreter lock (if it has been created and thread
  support is enabled) and set the thread state to <var>tstate</var>, which
  must not be <tt class="constant">NULL</tt>.  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.)
</dd></dl>

<P>
The following macros are normally used without a trailing semicolon;
look for example usage in the Python source distribution.

<P>
<dl><dt><b><tt id='l2h-849' xml:id='l2h-849' class="macro">Py_BEGIN_ALLOW_THREADS</tt></b></dt>
<dd>
  This macro expands to
  "<tt class="samp">{ PyThreadState *_save; _save = PyEval_SaveThread();</tt>".
  Note that it contains an opening brace; it must be matched with a
  following Py_END_ALLOW_THREADS macro.  See above for
  further discussion of this macro.  It is a no-op when thread support
  is disabled at compile time.
</dl>

<P>
<dl><dt><b><tt id='l2h-850' xml:id='l2h-850' class="macro">Py_END_ALLOW_THREADS</tt></b></dt>
<dd>
  This macro expands to "<tt class="samp">PyEval_RestoreThread(_save); }</tt>".
  Note that it contains a closing brace; it must be matched with an
  earlier Py_BEGIN_ALLOW_THREADS macro.  See above for
  further discussion of this macro.  It is a no-op when thread support
  is disabled at compile time.
</dl>

<P>
<dl><dt><b><tt id='l2h-851' xml:id='l2h-851' class="macro">Py_BLOCK_THREADS</tt></b></dt>
<dd>
  This macro expands to "<tt class="samp">PyEval_RestoreThread(_save);</tt>": it is
  equivalent to Py_END_ALLOW_THREADS without the
  closing brace.  It is a no-op when thread support is disabled at
  compile time.
</dl>

<P>
<dl><dt><b><tt id='l2h-852' xml:id='l2h-852' class="macro">Py_UNBLOCK_THREADS</tt></b></dt>
<dd>
  This macro expands to "<tt class="samp">_save = PyEval_SaveThread();</tt>": it is
  equivalent to Py_BEGIN_ALLOW_THREADS without the
  opening brace and variable declaration.  It is a no-op when thread
  support is disabled at compile time.
</dl>

<P>
All of the following functions are only available when thread support
is enabled at compile time, and must be called only when the
interpreter lock has been created.

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>PyInterpreterState*&nbsp;<b><tt id='l2h-853' xml:id='l2h-853' class="cfunction">PyInterpreterState_New</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
  Create a new interpreter state object.  The interpreter lock need
  not be held, but may be held if it is necessary to serialize calls
  to this function.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-854' xml:id='l2h-854' class="cfunction">PyInterpreterState_Clear</tt></b>(</nobr></td><td>PyInterpreterState *<var>interp</var>)</td></tr></table></dt>
<dd>
  Reset all information in an interpreter state object.  The
  interpreter lock must be held.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-855' xml:id='l2h-855' class="cfunction">PyInterpreterState_Delete</tt></b>(</nobr></td><td>PyInterpreterState *<var>interp</var>)</td></tr></table></dt>
<dd>
  Destroy an interpreter state object.  The interpreter lock need not
  be held.  The interpreter state must have been reset with a previous
  call to <tt class="cfunction">PyInterpreterState_Clear()</tt>.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>PyThreadState*&nbsp;<b><tt id='l2h-856' xml:id='l2h-856' class="cfunction">PyThreadState_New</tt></b>(</nobr></td><td>PyInterpreterState *<var>interp</var>)</td></tr></table></dt>
<dd>
  Create a new thread state object belonging to the given interpreter
  object.  The interpreter lock need not be held, but may be held if
  it is necessary to serialize calls to this function.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-857' xml:id='l2h-857' class="cfunction">PyThreadState_Clear</tt></b>(</nobr></td><td>PyThreadState *<var>tstate</var>)</td></tr></table></dt>
<dd>
  Reset all information in a thread state object.  The interpreter lock
  must be held.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-858' xml:id='l2h-858' class="cfunction">PyThreadState_Delete</tt></b>(</nobr></td><td>PyThreadState *<var>tstate</var>)</td></tr></table></dt>
<dd>
  Destroy a thread state object.  The interpreter lock need not be
  held.  The thread state must have been reset with a previous call to
  <tt class="cfunction">PyThreadState_Clear()</tt>.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>PyThreadState*&nbsp;<b><tt id='l2h-859' xml:id='l2h-859' class="cfunction">PyThreadState_Get</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
  Return the current thread state.  The interpreter lock must be
  held.  When the current thread state is <tt class="constant">NULL</tt>, this issues a fatal
  error (so that the caller needn't check for <tt class="constant">NULL</tt>).
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>PyThreadState*&nbsp;<b><tt id='l2h-860' xml:id='l2h-860' class="cfunction">PyThreadState_Swap</tt></b>(</nobr></td><td>PyThreadState *<var>tstate</var>)</td></tr></table></dt>
<dd>
  Swap the current thread state with the thread state given by the
  argument <var>tstate</var>, which may be <tt class="constant">NULL</tt>.  The interpreter lock
  must be held.
</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>PyObject*&nbsp;<b><tt id='l2h-861' xml:id='l2h-861' class="cfunction">PyThreadState_GetDict</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
<div class="refcount-info">
  <span class="label">Return value:</span>
  <span class="value">Borrowed reference.</span>
</div>
  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 <tt class="constant">NULL</tt>, no exception has been raised and the
  caller should assume no current thread state is available.

<span class="versionnote">Changed in version 2.3:
Previously this could only be called when a current
  thread is active, and <tt class="constant">NULL</tt> meant that an exception was raised.</span>

</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>int&nbsp;<b><tt id='l2h-862' xml:id='l2h-862' class="cfunction">PyThreadState_SetAsyncExc</tt></b>(</nobr></td><td>long <var>id</var>, PyObject *<var>exc</var>)</td></tr></table></dt>
<dd>
  Asynchronously raise an exception in a thread.
  The <var>id</var> argument is the thread id of the target thread;
  <var>exc</var> is the exception object to be raised.
  This function does not steal any references to <var>exc</var>.
  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; if it returns a number
  greater than one, you're in trouble, and you should call it again
  with <var>exc</var> set to <tt class="constant">NULL</tt> to revert the effect.
  This raises no exceptions.

<span class="versionnote">New in version 2.3.</span>

</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>PyGILState_STATE&nbsp;<b><tt id='l2h-863' xml:id='l2h-863' class="cfunction">PyGILState_Ensure</tt></b>(</nobr></td><td>)</td></tr></table></dt>
<dd>
Ensure that the current thread is ready to call the Python
C API regardless of the current state of Python, or of its
thread lock.  This may be called as many times as desired
by a thread as long as each call is matched with a call to
<tt class="cfunction">PyGILState_Release()</tt>.
In general, other thread-related APIs may
be used between <tt class="cfunction">PyGILState_Ensure()</tt> and <tt class="cfunction">PyGILState_Release()</tt> calls as long as the
thread state is restored to its previous state before the Release().
For example, normal usage of the Py_BEGIN_ALLOW_THREADS
and Py_END_ALLOW_THREADS macros is acceptable.

<P>
The return value is an opaque "handle" to the thread state when
<tt class="cfunction">PyGILState_Acquire()</tt> was called, and must be passed to
<tt class="cfunction">PyGILState_Release()</tt> 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
<tt class="cfunction">PyGILState_Ensure</tt> must save the handle for its call to
<tt class="cfunction">PyGILState_Release</tt>.

<P>
When the function returns, the current thread will hold the GIL.
Failure is a fatal error.

<span class="versionnote">New in version 2.3.</span>

</dd></dl>

<P>
<dl><dt><table cellpadding="0" cellspacing="0"><tr valign="baseline"><td><nobr>void&nbsp;<b><tt id='l2h-864' xml:id='l2h-864' class="cfunction">PyGILState_Release</tt></b>(</nobr></td><td>PyGILState_STATE)</td></tr></table></dt>
<dd>
Release any resources previously acquired.  After this call, Python's
state will be the same as it was prior to the corresponding
<tt class="cfunction">PyGILState_Ensure</tt> call (but generally this state will be
unknown to the caller, hence the use of the GILState API.)

<P>
Every call to <tt class="cfunction">PyGILState_Ensure()</tt> must be matched by a call to
<tt class="cfunction">PyGILState_Release()</tt> on the same thread.

<span class="versionnote">New in version 2.3.</span>

</dd></dl>

<P>

<DIV CLASS="navigation">
<div class='online-navigation'>
<p></p><hr />
<table align="center" width="100%" cellpadding="0" cellspacing="2">
<tr>
<td class='online-navigation'><a rel="prev" title="8. Initialization, Finalization, and"
  href="initialization.html"><img src='../icons/previous.png'
  border='0' height='32'  alt='Previous Page' width='32' /></A></td>
<td class='online-navigation'><a rel="parent" title="8. Initialization, Finalization, and"
  href="initialization.html"><img src='../icons/up.png'
  border='0' height='32'  alt='Up One Level' width='32' /></A></td>
<td class='online-navigation'><a rel="next" title="8.2 Profiling and Tracing"
  href="profiling.html"><img src='../icons/next.png'
  border='0' height='32'  alt='Next Page' width='32' /></A></td>
<td align="center" width="100%">Python/C API Reference Manual</td>
<td class='online-navigation'><a rel="contents" title="Table of Contents"
  href="contents.html"><img src='../icons/contents.png'
  border='0' height='32'  alt='Contents' width='32' /></A></td>
<td class='online-navigation'><img src='../icons/blank.png'
  border='0' height='32'  alt='' width='32' /></td>
<td class='online-navigation'><a rel="index" title="Index"
  href="genindex.html"><img src='../icons/index.png'
  border='0' height='32'  alt='Index' width='32' /></A></td>
</tr></table>
<div class='online-navigation'>
<b class="navlabel">Previous:</b>
<a class="sectref" rel="prev" href="initialization.html">8. Initialization, Finalization, and</A>
<b class="navlabel">Up:</b>
<a class="sectref" rel="parent" href="initialization.html">8. Initialization, Finalization, and</A>
<b class="navlabel">Next:</b>
<a class="sectref" rel="next" href="profiling.html">8.2 Profiling and Tracing</A>
</div>
</div>
<hr />
<span class="release-info">Release 2.4.2, documentation updated on 28 September 2005.</span>
</DIV>
<!--End of Navigation Panel-->
<ADDRESS>
See <i><a href="about.html">About this document...</a></i> for information on suggesting changes.
</ADDRESS>
</BODY>
</HTML>