2
0
mirror of https://github.com/boostorg/thread.git synced 2026-02-03 09:42:16 +00:00

Compare commits

...

5 Commits

Author SHA1 Message Date
nobody
4db2e35a33 This commit was manufactured by cvs2svn to create tag
'Version_1_25_0'.

[SVN r11317]
2001-10-01 16:59:12 +00:00
Beman Dawes
ca2cf20cf3 1.25.0 Final runup
[SVN r11315]
2001-10-01 15:54:23 +00:00
Beman Dawes
b3acba1d2d commit after merging thread-initial into main trunk
[SVN r11258]
2001-09-25 19:20:11 +00:00
Beman Dawes
6d2731c463 commit after merging thread-initial into main trunk
[SVN r11250]
2001-09-25 16:57:26 +00:00
Beman Dawes
b7f8f8867c merge branch "thread-initial" into main trunk
[SVN r11243]
2001-09-25 00:04:45 +00:00
57 changed files with 9922 additions and 0 deletions

40
build/Jamfile Normal file
View File

@@ -0,0 +1,40 @@
# (C) Copyright William E. Kempf 2001. Permission to copy, use, modify, sell and
# distribute this software is granted provided this copyright notice appears
# in all copies. This software is provided "as is" without express or implied
# warranty, and with no claim as to its suitability for any purpose.
#
# Boost.Threads build and test Jamfile
#
# Declares the following targets:
# 1. libboost_thread, a static link library.
# 1a. On Win32, a dynamic link library libboost_threadmon,
# which must be used in conjunction with libboost_thread.
# declare the location of this subproject relative to the root
subproject libs/thread/build ;
#######################
#
# Declare the Boost.Threads static link library.
#
# For Win32 we need to build a special DLL, libboost_threadmon, to handle
# TSS destruction.
if $(NT)
{
dll libboost_threadmon : ../src/threadmon.cpp
# requirements
: <threading>multi
: debug release ;
}
# Base names of the source files for libboost_thread
CPP_SOURCES =
condition mutex recursive_mutex semaphore thread tss xtime once ;
lib libboost_thread : ../src/$(CPP_SOURCES).cpp
# requirements
: <include>$(BOOST_ROOT)
<threading>multi
: debug release ;

72
doc/acknowledgements.html Normal file
View File

@@ -0,0 +1,72 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
<meta name="GENERATOR" content="Microsoft FrontPage 4.0">
<meta name="ProgId" content="FrontPage.Editor.Document">
<title>Boost.Threads Acknowledgements</title>
</head>
<body bgcolor="#FFFFFF" text="#000000">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Acknowledgements</h2>
</td>
</tr>
</table>
<h2>Acknowledgments</h2>
<p><a href="../../../people/william_kempf.htm">William E. Kempf</a> was the
architect, designer, and implementor of <b>Boost.Threads</b>.</p>
<p>Important contributions were also made by Jeremy Siek (lots of input on
the design and on the implementation), Alexander Terekhov (lots of input on
the Win32 implementation, especially in regards to boost::condition, as well
as a lot of explanation of POSIX behavior), Greg Colvin (lots of input on the
design), and Paul Mclachlan, Thomas Matelich and Iain Hanson (for help in
trying to get the build to work on other platforms).</p>
<p>The documentation was written by William E. Kempf. Beman Dawes provided
additional documentation material and editing.</p>
<p>Discussions on the boost.org mailing list were essential in the development
of <b>Boost.Threads</b>. As of August 1, 2001, participants included Alan Griffiths,
Albrecht Fritzsche, Aleksey Gurtovoy, Alexander Terekhov, Andrew Green, Andy Sawyer,
Asger Alstrup Nielsen, Beman Dawes, Bill Klein, Bill Rutiser, Bill Wade, Branko
Èibej, Brent Verner, Craig Henderson, Csaba Szepesvari, Dale Peakall, Damian
Dixon, Dan Nuffer, Darryl Green, Daryle Walker, David Abrahams, David Allan
Finch, Dejan Jelovic, Dietmar Kuehl, Doug Gregor, Douglas Gregor, Duncan Harris,
Ed Brey, Eric Swanson, Eugene Karpachov, Fabrice Truillot, Frank
Gerlach, Gary Powell, Gernot Neppert, Geurt Vos, Ghazi Ramadan, Greg Colvin,
Gregory Seidman, HYS, Iain Hanson, Ian Bruntlett, J Panzer, Jeff Garland, Jeff
Paquette, Jens Maurer, Jeremy Siek, Jesse Jones, Joe Gottman, John (EBo) David,
John Bandela, John Maddock, John Max Skaller, John Panzer, Jon Jagger , Karl
Nelson, Kevlin Henney, KG Chandrasekhar, Levente Farkas, Lie-Quan Lee, Lois
Goldthwaite, Luis Pedro Coelho, Marc Girod, Mark A. Borgerding, Mark Rodgers,
Marshall Clow, Matthew Austern, Matthew Hurd, Michael D. Crawford, Michael H.
Cox , Mike Haller, Miki Jovanovic, Nathan Myers, Paul Moore, Pavel Cisler, Peter
Dimov, Petr Kocmid, Philip Nash, Rainer Deyke, Reid Sweatman, Ross Smith, Scott
McCaskill, Shalom Reich , Steve Cleary, Steven Kirk, Thomas Holenstein, Thomas
Matelich, Trevor Perrin, Valentin Bonnard, Vesa Karvonen, Wayne Miller, and
William Kempf.</p>
<p>Apologies for anyone inadvertently missed.</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p>©<i> Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001</i></p>
</body>
</html>

145
doc/bibliography.html Normal file
View File

@@ -0,0 +1,145 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
<meta name="GENERATOR" content="Microsoft FrontPage 4.0">
<meta name="ProgId" content="FrontPage.Editor.Document">
<title>Boost.Threads Bibliography</title>
</head>
<body bgcolor="#FFFFFF" text="#000000">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Bibliography</h2>
</td>
</tr>
</table>
<h2>Bibliography</h2>
<table border="0" cellpadding="5" width="777">
<tr>
<td width="102" valign="top" align="left"><b>[<a name="Andrews-83">Andrews
83</a>]</b></td>
<td width="645">
Gregory R. Andrews, Fred B. Schneider, <cite>Concepts and Notations for Concurrent
Programming</cite>, ACM Computing Surveys, Vol. 15, No. 1, March, 1983. <a href="http://www.acm.org/pubs/citations/journals/surveys/1983-15-1/p3-andrews/">http://www.acm.org/pubs/citations/journals/surveys/1983-15-1/p3-andrews/</a>
<p>Good general background reading.&nbsp; Includes descriptions of Path
Expressions, Message Passing, and Remote Procedure Call in addition to the
basics.&nbsp;
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"><b>[<a name="Boost">Boost</a>]</b></td>
<td width="645">
The <cite> Boost</cite> world-wide web site.&nbsp; <a href="http://www.boost.org">http://www.boost.org</a>
<p>Boost.Threads is one of many Boost libraries.&nbsp; The Boost web site
includes a great deal of documentation and general information which applies to
all Boost libraries. Current copies of the libraries including documentation and
test programs may be downloaded from the web site.
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"><b>[<a name="Brinch-Hansen-73">Brinch
Hansen 73</a>]</b></td>
<td width="645">
Per Brinch Hansen, <cite>Concurrent Programming Concepts</cite>, ACM Computing
Surveys, Vol. 5, No. 4, December, 1973. <a href="http://www.acm.org/pubs/articles/journals/surveys/1973-5-4/p223-hansen/p223-hansen.pdf">http://www.acm.org/pubs/articles/journals/surveys/1973-5-4/p223-hansen/</a>
<p>&quot;This paper describes the evolution of language features for
multiprogramming from event queues and semaphores to critical regions and
monitors.&quot; Includes analysis of&nbsp; why <i>events</i> are considered
error-prone. Also noteworthy because of an introductory quotation from
Christopher Alexander; Brinch Hansen was years ahead of others in recognizing
pattern concepts applied to software too.
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"><b>]<a name="Butenhof-97">Butenhof
97</a>]</b></td>
<td width="645">
<p> David R. Butenhof, <cite>Programming with
POSIX Threads</cite>, Addison-Wesley 1997, ISBN 0-201-63392-2&nbsp; <a href="http://cseng.aw.com/book/0,3828,0201633922,00.html">http://cseng.aw.com/book/0,3828,0201633922,00.html</a></p>
<p>This is a very readable explanation of threads and how to use them.&nbsp; Many
of the insights given apply to all multi-threaded programming, not just POSIX
Threads.</p>
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"><b>[<a name="Hoare-74">Hoare 74</a>]</b></td>
<td width="645">
<p>C.A.R Hoare, <cite> Monitors: An Operating System Structuring Concept</cite>,
Communications of the ACM, Vol. 17, No. 10. October
1974, pp. 549-557&nbsp; <a href="http://www.acm.org/classics/feb96/">http://www.acm.org/classics/feb96/ </a></p>
<p>Hoare and Brinch Hansen's work on Monitors is the basis for reliable
multi-threading patterns. This is one of the most often referenced papers in
all of computer science, and with good reason.</p>
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"><b>[<a name="ISO-98">ISO 98</a>]</b></td>
<td width="645">
<p>ISO/IEC 14882:1998(E) <cite> Programming Language C++</cite>&nbsp; <a href="http://www.ansi.org">http://www.ansi.org</a></p>
<p>This is the official C++ Standards
document. Available from the ANSI (American
National Standards Institute) Electronic Standards Store.</p>
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"><b>[<a name="McDowell-89">McDowell
89</a>]</b></td>
<td width="645">
Charles E McDowell, David P. Helmbold, <cite>Debugging Concurrent Programs</cite>,
ACM Computing Surveys, Vol. 21, No. 2, December, 1989. <a href="http://www.acm.org/pubs/citations/journals/surveys/1989-21-4/p593-mcdowell/">http://www.acm.org/pubs/citations/journals/surveys/1989-21-4/p593-mcdowell/</a>
<p>Identifies many of the unique failure modes and debugging difficulties
associated with concurrent programs.
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"> <b>[<a name="Schmidt">Schmidt</a>]&nbsp;</b></td>
<td width="645">
<p> Douglas C. Schmidt and Irfan Pyarali, <cite>Strategies for
Implementing POSIX Condition Variables on Win32</cite>,&nbsp; Department of Computer Science, Washington University, St. Louis, Missouri.&nbsp;
<a href="http://www.cs.wustl.edu/~schmidt/win32-cv-1.html">http://www.cs.wustl.edu/~schmidt/win32-cv-1.html</a></p>
<p>Rationale for understanding Boost.Threads condition variables. Note that Alexander Terekhov found some bugs in
the implementation given in this article, so pthreads-win32 and Boost.Threads
are even more complicated yet.</p>
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"> <b>[<a name="Schmidt-00">Schmidt
00</a>]&nbsp;</b></td>
<td width="645">
<p> Douglas C. Schmidt, Michael Stal, Hans Rohnert and Frank Buschmann, <cite>Pattern-Oriented Software Architecture Volume 2 - Patterns for
Concurrent and Networked Objects</cite>,&nbsp; Wiley 2000, ISBN 0-471-60695-2&nbsp; <a href="http://www.wiley.com/Corporate/Website/Objects/Products/0,9049,104671,00.html">http://www.wiley.com/Corporate/Website/Objects/Products/0,9049,104671,00.html</a></p>
<p>This is a very good explanation of how to apply several patterns useful for concurrent programming.
Among the patterns documented is the Monitor Pattern mentioned frequently in the <b>Boost.Threads</b>
documentation.</p>
</td>
</tr>
<tr>
<td width="102" valign="top" align="left"> <b>[<a name="Stroustrup-00">Stroustrup
00</a>]</b></td>
<td width="645">
Bjarne Stroustrup, <cite> The C++ Programming Language</cite>, Special Edition, Addison-Wesley
2000, ISBN 0-201-70073-5 <a href="http://cseng.aw.com/book/0,3828,0201700735,00.html">http://cseng.aw.com/book/0,3828,0201700735,00.html</a>
<p>The first book a C++ programmer should own.&nbsp; Note that the 3rd edition
(and subsequent editions like the Special Edition) has been rewritten to cover
the ISO standard language and library.
</td>
</tr>
</table>
<p>Note: The URL's above are provided in plain text form so that they will be visible
on printed copies of this document.</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %b %Y" startspan -->01 Oct 2001<!--webbot bot="Timestamp" endspan i-checksum="14990" -->
</p>
<p>© Copyright Beman Dawes, 2001</p>
</body>
</html>

128
doc/call_once.html Normal file
View File

@@ -0,0 +1,128 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, Boost.Threads, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, call_once</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">call_once</h2>
</td>
</tr>
</table>
<hr>
<p><A href="#Introduction">Introduction</A><br>
<A href="#Header">Header</A><br>
<A href="#Synopsis">Synopsis</A><br>
<A href="#Members">Members</A><br>
<A href="#Example">Example</A></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <code>call_once</code> routine and <code>once_flag</code> type can be used to
run a routine exactly once. This can be used to initialize data in a
<a href="definitions.html#Thread-safe">thread-safe</a> manner.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/once.hpp">&lt;boost/thread/once.hpp&gt;</a>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost {
typedef <i>[implementation defined]</i> once_flag;
const once_flag once_init = <i>[implementation defined]</i>;
void call_once(void (*func)(), once_flag&amp; flag);
} // namespace boost
</pre>
<h2><a name="Reference">Reference</a></h2>
<hr>
<h3>once_flag</h3>
<p>This implementation defined type is used as a flag to insure a routine is called only once.
Instances of this type should be statically initialized to <code>once_init</code>.</p>
<hr>
<h3>once_init</h3>
<p>This is a constant value used to initialize <code>once_flag</code> instances
to indicate that the logically associated routine has not been run yet.</p>
<hr>
<h3>call_once</h3>
<pre>
void call_once(void (*func)(), once_flag&amp; flag);
</pre>
<p><b>Requires:</b> The function <code>func</code> shall not throw exceptions.</p>
<p><b>Effects:</b> As if (in an atomic fashion)</p>
<code>
&nbsp;&nbsp;&nbsp;if (flag == once_init)<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;func();
</code>
<p><b>Postcondition:</b> <code>flag</code> != <code>once_init</code></p>
<hr>
<h2><a name="Example">Example Usage</a></h2>
<pre>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
#include <a href="../../../boost/thread/tss.hpp">&lt;boost/thread/once.hpp&gt;</a>
#include &lt;cassert&gt;
int value=0;
boost::once_flag once = boost::once_init;
void init()
{
++value;
}
void thread_proc()
{
boost::call_once(&amp;init, once);
}
int main(int argc, char* argv[])
{
boost::thread_group threads;
for (int i=0; i&lt;5; ++i)
threads.create_thread(&amp;thread_proc);
threads.join_all();
assert(value == 1);
}
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

333
doc/condition.html Normal file
View File

@@ -0,0 +1,333 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, condition</title>
</head>
<body bgcolor="#ffffff" link="#0000ff" vlink="#800080" text="#000000">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><IMG height=86 alt="C++ Boost" src="../../../c++boost.gif" width=277></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">condition</h2>
</td>
</tr>
</table>
<hr>
<p><A href="#Introduction">Introduction</a><br>
<A href="#Header">Header</a><br>
<A href="#Synopsis">Synopsis</a><br>
<A href="#Members">Members</a><br>
<A href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>An object of class <code>condition</code> is a synchronization primitive used to
cause a thread to wait until a particular shared-data condition (or time) is met.
A <code>condition</code> object is always used in conjunction with a mutex
object modeling a <a href="mutex_concept.html">Mutex Concept</a>. The mutex must be locked prior to waiting on the
<code>condition</code>, which is ensured by passing a lock object modeling a <a href="lock_concept.html">Lock
Concept</a> to the <code>condition</code> object's <code>wait</code> functions. While the thread is waiting on the <code>condition</code>
object,
the mutex associated with the lock is unlocked. When the thread returns
from a call to one of the <code>condition</code> object's <code>wait</code> functions,
the mutex is again locked. The tricky lock/unlock/lock sequence is performed
automatically by the <code>condition</code> object's <code>wait</code>
functions.</p>
<p>The <code>condition</code> type is often used to implement the <i>Monitor Object</i>
and other important patterns. See <A href="bibliography.html#Schmidt-00">[Schmidt-00]</a>
and <A href="bibliography.html#Hoare-74">[Hoare 74]</a>. Monitors are one of the most
important patterns for creating reliable multithreaded programs.</p>
<p>See <A href="definitions.html">Formal Definitions</a> for definitions of thread
states <A href="definitions.html#state">blocked</a> and
<A href="definitions.html#state">ready</a>. Note that "waiting" is a synonym
for blocked.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <A href="../../../boost/thread/condition.hpp">&lt;boost/thread/condition.hpp&gt;</a>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost {
class condition : private <A href="../../utility/utility.htm#Class noncopyable">boost::noncopyable</a> // Exposition only.
// Class condition meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
condition();
~condition();
void notify_one();
void notify_all();
template &lt;typename <a href="scoped_lock.html">ScopedLock</a>&gt;
void wait(<a href="scoped_lock.html">ScopedLock</a>&amp; lock);
template &lt;typename <a href="scoped_lock.html">ScopedLock</a>, typename <A href="http://www.sgi.com/tech/stl/Predicate.html">Predicate</A>&gt;
void wait(<a href="scoped_lock.html">ScopedLock</a>&amp; lock, <A href="http://www.sgi.com/tech/stl/Predicate.html">Predicate</A> pred);
template &lt;typename <a href="scoped_lock.html">ScopedLock</a>&gt;
bool timed_wait(<a href="scoped_lock.html">ScopedLock</a>&amp; lock, const xtime&amp; xt);
template &lt;typename <a href="scoped_lock.html">ScopedLock</a>, typename <A href="http://www.sgi.com/tech/stl/Predicate.html">Predicate</A>&gt;
bool timed_wait(<a href="scoped_lock.html">ScopedLock</a>&amp; lock, const xtime&amp; xt, <A href="http://www.sgi.com/tech/stl/Predicate.html">Predicate</A> pred);
};
} // namespace boost
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
condition();
</pre>
<p><b>Effects:</b> Constructs a <code>condition</code>.</p>
<hr>
<h3>Destructor</h3>
<pre>
~condition();
</pre>
<p><b>Effects:</b> Destroys <code>*this</code>.</p>
<hr>
<h3>notify_one</h3>
<pre>
void notify_one();
</pre>
<p><b>Effects:</b> If there is a thread waiting on <code>*this</code>, change
that thread's state to ready. Otherwise there is no effect.</p>
<p><b>Notes:</b> If more that one thread is waiting on the condition, it is
unspecified which is made ready.</p>
<hr>
<h3>notify_all</h3>
<pre>
void notify_all();
</pre>
<p><b>Effects:</b> Change the state of all threads waiting on <code>*this</code>
to ready. If there are no waiting threads, <code>notify_all()</code> has no effect.</p>
<hr>
<h3>wait</h3>
<pre>
template &lt;typename ScopedLock&gt;
void wait(ScopedLock&amp; lock);
</pre>
<p><b>Requires:</b> ScopedLock meets the
<A href="lock_concept.html#ScopedLock">ScopedLock</a> requirements.</p>
<p><b>Effects:</b> Releases the lock on the <A href="mutex_concept.html">mutex model</a>
associated with <code>lock</code>, blocks the current thread of execution until readied
by a call to <code>this-&gt;notify_one()</code> or <code>this-&gt;notify_all()</code>,
and then reacquires the lock. All effects occur in an atomic fashion.</p>
<p><b>Throws:</b> <code><A href="lock_error.html">lock_error</a></code>
if <code>!lock.locked()</code></p>
<p><b>Danger:</b> This version should always be used within a loop checking that the
state logically associated with the <code>condition</code> has become true. Without
the loop, race conditions can ensue due to possible "spurious wake ups". The second
version encapsulates this loop idiom internally and is generally the preferred method.</p>
<pre>
template &lt;typename ScopedLock, typename Pr&gt;
void wait(ScopedLock&amp; lock, Pr pred);
</pre>
<p><b>Requires:</b> ScopedLock meets the
<A href="lock_concept.html#ScopedLock">ScopedLock</a> requirements, return from
<code>pred()</code> convertible to bool.</p>
<p><b>Effects:</b> As if:</p>
<code>
&nbsp;&nbsp;&nbsp;while (!pred()) wait(lock)
</code>
<p><b>Throws:</b> <code><A href="lock_error.html">lock_error</a></code> if
<code>!lock.locked()</code></p>
<hr>
<h3>timed_wait</h3>
<pre>
template &lt;typename ScopedTimedLock&gt;
bool timed_wait(ScopedTimedLock&amp; lock, const <a href="xtime.html">xtime</a>&amp; xt);
</pre>
<p><b>Requires:</b> ScopedTimeLock meets the
<A href="lock_concept.html#ScopedTimedLock">ScopedTimedLock</a> requirements.</p>
<p><b>Effects:</b> Releases the lock on the <A href="mutex_concept.html">mutex model</a>
associated with the <code>lock</code>, blocks the current thread of execution until
readied by a call to <code>this-&gt;notify_one()</code> or
<code>this-&gt;notify_all()</code>, or until <code>xt</code>, and then reacquires the
lock. All effects occur in an atomic fashion.</p>
<p><b>Throws:</b> <code><A href="lock_error.html">lock_error</a></code> if
<code>!lock.locked()</code></p>
<p><b>Danger:</b> This version should always be used within a loop checking that the
state logically associated with the <code>condition</code> has become true. Without
the loop, race conditions can ensue due to "spurious wake ups". The second version
encapsulates this loop idiom internally and is generally the preferred method.</p>
<p><b>Returns:</b> <code>false</code> if <code>xt</code> is reached, otherwise
<code>true</code>.</p>
<pre>
template &lt;typename ScopedTimedLock, typename Pr&gt;
bool timed_wait(ScopedTimedLock&amp; lock, const <a href="xtime.html">xtime</a>&amp; xt, Pr pred);
</pre>
<p><b>Requires: </b>ScopedTimeLock meets the
<A href="lock_concept.html#ScopedTimedLock">ScopedTimedLock</a> requirements,
return from <code>pred()</code> convertible to bool.</p>
<p><b>Effects: </b>As if:</p>
<code>
&nbsp;&nbsp;&nbsp;while (!pred())<br>
&nbsp;&nbsp;&nbsp;{<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;if (!timed_wait(lock, xt))<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;return false;<br>
&nbsp;&nbsp;&nbsp;}
</code>
<p><b>Throws:</b> <code><A href="lock_error.html">lock_error</a></code> if
<code>!lock.locked()</code></p>
<p><b>Returns:</b> <code>false</code> if <code>xt</code> is reached, otherwise
<code>true</code>.</p>
<hr>
<h2><a name="Example">Example Usage</a></h2>
<pre>
#include &lt;iostream&gt;
#include &lt;vector&gt;
#include <A href="../../../boost/utility.hpp">&lt;boost/utility.hpp&gt;</a>
#include <A href="../../../boost/thread/condition.hpp">&lt;boost/thread/condition.hpp&gt;</a>
#include <A href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
class bounded_buffer : private boost::noncopyable
{
public:
typedef boost::mutex::scoped_lock lock;
bounded_buffer(int n) : begin(0), end(0), buffered(0), circular_buf(n) { }
void send (int m) {
lock lk(monitor);
while (buffered == circular_buf.size())
buffer_not_full.wait(lk);
circular_buf[end] = m;
end = (end+1) % circular_buf.size();
++buffered;
buffer_not_empty.notify_one();
}
int receive() {
lock lk(monitor);
while (buffered == 0)
buffer_not_empty.wait(lk);
int i = circular_buf[begin];
begin = (begin+1) % circular_buf.size();
--buffered;
buffer_not_full.notify_one();
return i;
}
private:
int begin, end, buffered;
std::vector&lt;int&gt; circular_buf;
boost::condition buffer_not_full, buffer_not_empty;
boost::mutex monitor;
};
bounded_buffer buf(2);
void sender() {
int n = 0;
while (n &lt; 100) {
buf.send(n);
std::cout &lt;&lt; &quot;sent: &quot; &lt;&lt; n &lt;&lt; std::endl;
++n;
}
buf.send(-1);
}
void receiver() {
int n;
do {
n = buf.receive();
std::cout &lt;&lt; &quot;received: &quot; &lt;&lt; n &lt;&lt; std::endl;
} while (n != -1); // -1 indicates end of buffer
}
int main(int, char*[])
{
boost::thread thrd1(&amp;sender);
boost::thread thrd2(&amp;receiver);
thrd1.join();
thrd2.join();
return 0;
}
</pre>
<p>Typical output (dependent on scheduling policies) is:</p>
<pre>
sent: 0
sent: 1
received: 0
received: 1
sent: 2
sent: 3
received: 2
received: 3
sent: 4
received: 4
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <A href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

103
doc/config.html Normal file
View File

@@ -0,0 +1,103 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<title>Boost.Threads, Configuration Information</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Configuration Information</h2>
</td>
</tr>
</table>
<hr>
<p><b>Boost.Threads</b> uses several configuration macros in <a href="../../config/config.htm"> &lt;boost/config.hpp&gt;</a>.
These macros are documented here. Most of the macros are
of interest only to developers attempting to provide new implementations of <b>Boost.Threads</b>.
The one exception to this is BOOST_HAS_THREADS.</p>
<table cellspacing="10" width="100%">
<tr>
<td valign="top">
<b>Macro</b>
</td>
<td valign="top">
<b>Meaning</b>
</td>
</tr>
<tr>
<td valign="top">
BOOST_HAS_THREADS
</td>
<td valign="top">
Indicates that threading support is available. This means both that there is a
platform specific implementation for <b>Boost.Threads</b> and that threading
support has been enabled in a platform specific manner. For instance, on the
Win32 platform there's an implementation for <b>Boost.Threads</b> but unless
the program is compiled against one of the multi-threading runtimes
(often determined by the
compiler predefining the macro _MT) the
BOOST_HAS_THREADS macro remains undefined.
</td>
</tr>
<tr>
<td valign="top">
BOOST_HAS_WINTHREADS
</td>
<td valign="top">
Indicates that the platform has the Microsoft Win32 threading libraries,
and that they should be used
to implement <b>Boost.Threads</b>.
</td>
</tr>
<tr>
<td valign="top">
BOOST_HAS_PTHREADS
</td>
<td valign="top">
Indicates that the platform has the POSIX pthreads libraries, and that
they should be used
to implement <b>Boost.Threads</b>.
</td>
</tr>
<tr>
<td valign="top">
BOOST_HAS_FTIME
</td>
<td valign="top">
Indicates that the implementation should use GetSystemTimeAsFileTime() and
the FILETIME type to calculate the current time. This is an implementation
detail used by boost::detail::getcurtime().
</td>
</tr>
<tr>
<td valign="top">
BOOST_HAS_GETTTIMEOFDAY
</td>
<td valign="top">
Indicates that the implementation should use gettimeofday() to calculate the
current time. This is an implementation detail used by boost::detail::getcurtime().
</td>
</tr>
</table>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

244
doc/definitions.html Normal file
View File

@@ -0,0 +1,244 @@
<html>
<head>
<meta http-equiv="Content-Language" content="en-us">
<meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
<meta name="GENERATOR" content="Microsoft FrontPage 4.0">
<meta name="ProgId" content="FrontPage.Editor.Document">
<title>Boost.Threads Definitions</title>
</head>
<body bgcolor="#FFFFFF">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center"> Definitions</h2>
</td>
</tr>
</table>
<h2>Introduction</h2>
<p>The definitions are given in terms of the <a href="bibliography.html#ISO-98"> C++ Standard</a>.&nbsp; References to the standard
are in the form [1.2.3/4], which
represents the section number, with the paragraph number following the &quot;/&quot;.</p>
<p>Because the definitions are written in something akin to
&quot;standardese&quot;, they can be difficult to understand.&nbsp; The intent
isn't to confuse, but rather to clarify the additional requirements
Boost.Threads places on a C++ implementation as defined by the C++ Standard.</p>
<h2>Definitions</h2>
<h3>Thread</h3>
<p>Thread is short for &quot;thread of execution&quot;. A thread of execution is an execution environment [1.9/7] within the execution environment
of a C++ program [1.9]. The main() function [3.6.1] of the program is the
initial function of the initial thread. A program in a multi-threading
environment always has an initial thread even if the program explicitly creates
no additional threads.</p>
<p>Unless otherwise specified, each thread shares all aspects of its execution environment with
other threads in the program. Shared aspects of the execution environment
include, but are not limited to, the following:</p>
<ul>
<li>Static storage duration (static, extern) objects [3.7.1].</li>
</ul>
<ul>
<li>Dynamic storage duration (heap) objects [3.7.3].&nbsp; Thus each memory
allocation will return a unique addresses, regardless of the thread making
the allocation request.</li>
</ul>
<ul>
<li>Automatic storage duration (stack) objects [3.7.2] accessed via pointer or
reference from another thread.</li>
</ul>
<ul>
<li>Resources provided by the operating
system.&nbsp; For example, files.</li>
</ul>
<ul>
<li>The program itself.&nbsp; In other words, each thread is executing some
function of the same program, not a totally different program.</li>
</ul>
<p>Each thread has its own:</p>
<ul>
<li>Registers and current execution sequence (program counter) [1.9/5].</li>
</ul>
<ul>
<li>Automatic storage duration (stack) objects [3.7.2].</li>
</ul>
<h3><a name="Thread-safe">Thread-safe</a></h3>
<p>A program is thread-safe if it has no <a href="#Race condition">race
conditions</a>, does not <a href="#Deadlock">deadlock</a>, and has no <a href="#Priority failure">priority
failures</a>.</p>
<p>Note that thread-safety does not necessarily imply efficiency, and than while
some thread-safety violations can be determined statically at compile time, many
thread-safety errors can only only be detected at runtime.</p>
<h3>Thread <a name="State">State</a></h3>
<p>During the lifetime of a thread, it shall be in one of the following
states:</p>
<table border="1" cellpadding="5">
<tr>
<td><b>State</b></td>
<td><b>Description</b></td>
</tr>
<tr>
<td>Ready</td>
<td>Ready to run, but waiting for a processor.</td>
</tr>
<tr>
<td>Running</td>
<td>Currently executing on a processor. Zero or more threads may be running
at any time, with a maximum equal to the number of processors.&nbsp;</td>
</tr>
<tr>
<td>Blocked</td>
<td>Waiting for some resource other than a processor which is not currently
available, or for the completion of calls to library functions [1.9/6].
The term &quot;waiting&quot; is synonymous for &quot;blocked&quot;</td>
</tr>
<tr>
<td>Terminated</td>
<td>Finished execution but not yet detached or joined.</td>
</tr>
</table>
<p>Thread state transitions shall occur only as specified:</p>
<table border="1" cellpadding="5">
<tr>
<td><b>From</b></td>
<td><b>To</b></td>
<td><b>Cause</b></td>
</tr>
<tr>
<td>
<p align="left">[none]</td>
<td>Ready</td>
<td>Thread is created by a call to a library function.&nbsp; In the case of
the initial thread, creation is implicit and occurs during the startup of
the main() function [3.6.1].</td>
</tr>
<tr>
<td>Ready</td>
<td>Running</td>
<td>Processor becomes available.</td>
</tr>
<tr>
<td>Running</td>
<td>Ready</td>
<td>Thread preempted.</td>
</tr>
<tr>
<td>Running</td>
<td>Blocked</td>
<td>Thread calls a library function which waits for a resource or for the
completion of I/O.</td>
</tr>
<tr>
<td>Running</td>
<td>Terminated</td>
<td>Thread returns from its initial function, calls a thread termination
library function, or is cancelled by some other thread calling a thread
termination library function.</td>
</tr>
<tr>
<td>Blocked</td>
<td>Ready</td>
<td>The resource being waited for becomes available, or the blocking library
function completes.</td>
</tr>
<tr>
<td>Terminated</td>
<td>[none]</td>
<td>Thread is detached or joined by some other thread calling the
appropriate library function, or by program termination [3.6.3].</td>
</tr>
</table>
<p>[Note: if a suspend() function is added to the threading library, additional
transitions to the blocked state will have to be added to the above table.]</p>
<h3><a name="Race condition">Race condition</a></h3>
<p>A race condition is what occurs when multiple threads read and
write to the same memory without proper synchronization, resulting in an
incorrect value being read or written.&nbsp; The result of a race condition may
be a bit pattern which isn't even a valid value for the data type. A race
condition results in undefined behavior [1.3.12].</p>
<p>Race conditions can be prevented by serializing memory access
using the tools provided by Boost.Threads.&nbsp;</p>
<h3><a name="Deadlock">Deadlock</a></h3>
<p>Deadlock is an execution state where for some set of threads, each thread in
the set is blocked waiting for some action by one of the other threads in the
set. Since each is waiting on the others, none will ever become ready again.</p>
<h3><a name="Priority failure">Priority failure</a></h3>
<p>A priority failure (such as priority inversion or infinite overtaking) occurs
when threads executed in such a sequence that required work is not performed in
time to be useful.</p>
<h2>Memory visibility between threads</h2>
<p>An address [1.7] shall always point to the same memory byte, regardless of the
thread or processor dereferencing the address.</p>
<p>An object [1.8, 1.9] is accessible from multiple threads if it is of
static storage duration (static, extern) [3.7.1], or if a pointer or reference to
it is explicitly or
implicitly dereferenced in multiple threads.</p>
<p>For an object accessible from multiple threads, the value of the object
accessed from one thread may be indeterminate or different than the value
accessed from another thread, except under the conditions specified in the following
table.&nbsp;&nbsp; For the same row of the table, the value of an object
accessible at the indicated sequence point in thread A will be determinate and the
same if accessed at or after the indicated sequence point in thread B, provided
the object is not otherwise modified. In the table, the
&quot;sequence point at a call&quot; is the sequence point after the evaluation
of all function arguments [1.9/17], while the &quot;sequence point after a
call&quot; is the sequence point after the copying of the returned
value...&quot; [1.9/17].</p>
<table border="1" cellpadding="5">
<tr>
<td align="center"><b>Thread A</b></td>
<td align="center"><b>Thread B</b></td>
</tr>
<tr>
<td>The sequence point at a call to a library thread-creation
function.&nbsp;</td>
<td>The first sequence point of the initial function in the new thread
created by the Thread A call.</td>
</tr>
<tr>
<td>The sequence point at a call to a library function which locks a mutex,
directly or by waiting for a condition variable.</td>
<td>The sequence point after a call to a library function which unlocks the
same mutex.</td>
</tr>
<tr>
<td>The last sequence point before thread termination.</td>
<td>The sequence point after a call to a library function which joins the
terminated thread.</td>
</tr>
<tr>
<td>The sequence point at a call to a library function which signals or
broadcasts a condition variable.</td>
<td>The sequence point after the call to the library function which was
waiting on that same condition variable or signal.</td>
</tr>
</table>
<p>The architecture of the execution environment and the observable behavior of
the abstract machine [1.9] shall be the same on all processors.</p>
<p>The latitude granted by the C++ standard for an implementation to alter the
definition of observable behavior of the abstract machine to include additional library I/O
functions [1.9/6] is extended to include threading library functions.</p>
<p>When an exception is thrown and there is no matching exception handler in the
same thread, behavior is undefined.&nbsp; The preferred behavior is the same as when there is no matching exception handler
in a program [15.3/9].&nbsp; That is, terminate() is called, and it is implementation defined
whether or not the stack is unwound.</p>
<h2><a name="Acknowledgements">Acknowledgements</a></h2>
<p>This document has been much improved by the incorporation of comments from
William Kempf.</p>
<p>The visibility rules are based on <a href="bibliography.html#Butenhof-97">[Butenhof
97]</a>.&nbsp;</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %b %Y" startspan -->01 Oct 2001<!--webbot bot="Timestamp" endspan i-checksum="14990" -->
</p>
<p>© Copyright Beman Dawes, 2001</p>
<p>&nbsp;</p>
<p>&nbsp;</p>
<p>&nbsp;</p>
</body>
</html>

174
doc/faq.html Normal file
View File

@@ -0,0 +1,174 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, FAQ</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Frequently Asked Questions</h2>
</td>
</tr>
</table>
<hr>
<h2>1. Are lock objects <a href="definitions.html#Thread-safe">thread-safe</a>?</h2>
<p><b>No!</b> Lock objects are not meant to be shared between threads. They are meant to
be short lived objects created on automatic storage within a code block. Any other usage
is just likely to lead to errors and won't really be of actual benefit any way. Share
<a href="mutex_concept.html">mutexes</a>, not locks. For more information see the
<a href="rationale.html#lock_objects">rationale</a> behind the design for lock objects.</p>
<h2>2a. Why was Boost.Threads modeled after (specific library name)?</h2>
<p>It wasn't. Boost.Threads was designed from scratch. Extensive design
discussions involved numerous people representing a wide range of experience across
many platforms. To ensure portability, the initial implements were done in
parallel using POSIX Threads and theWin32 threading API. But the Boost.Threads
design is very much in the spirit of C++, and thus doesn't model such C based
APIs.</p>
<h2>2b. Why wasn't Boost.Threads modeled after (specific library name)?</h2>
<p>Existing C++ libraries either seemed dangerous (often failing to take
advantage of prior art to reduce errors) or had excessive dependencies on
library components unrelated to threading. Existing C libraries couldn't meet
our C++ requirements, and were also missing certain features. For
instance, POSIX threads doesn't support a maximum value for semaphores.
The WIN32 thread API lacks condition variables, even though these are critical
for the important Monitor pattern <a href="bibliography.html#Schmidt-00">[Schmidt
00]</a>.</p>
<h2>3. Why do <a href="mutex_concept.html"> Mutexes</a> have noncopyable semantics?</h2>
<p>To ensure that <a href="definitions.html#Deadlock"> deadlocks</a> don't occur. The only logical form of copy would be to
use some sort of shallow copy semantics in which multiple mutex objects could refer
to the same mutex state. This means that if ObjA has a mutex object as part of its state
and ObjB is copy constructed from it, then when ObjB::foo() locks the mutex it has effectively
locked ObjA as well. This behavior can result in deadlock. Other
copy semantics result in similar problems (if you think you can prove this to
be wrong then supply us with an alternative and we'll reconsider).</p>
<h2>4. How can you prevent <a href="definitions.html#Deadlock"> deadlock</a> from occurring when a thread must lock multiple
mutexes?</h2>
<p>Always lock them in the same order. One easy way of doing this is to use
each mutex's address to determine the order in which they are locked. A future
Boost.Threads concept may wrap this pattern up in a reusable class.</p>
<h2>5. Don't noncopyable <a href="mutex_concept.html"> mutex</a> semantics mean that a
class with a mutex member will be noncopyable as well?</h2>
<p>No, but what it does mean is that the compiler can't generate a copy constructor
and assignment operator, so they will have to be coded explicitly. This is a
<b>good thing</b>, however, since the compiler generated operations would not
be <a href="definitions.html#Thread-safe">thread-safe</a>. The following is a
simple example of a class with copyable semantics and internal synchronization through
a mutex member.</p>
<pre>
class counter
{
public:
// Doesn't need synchronization since there can be no references to *this
// until after it's constructed!
explicit counter(int initial_value)
: m_value(initial_value)
{
}
// We only need to syncronize other for the same reason we don't have to
// synchronize on construction!
counter(const counter&amp; other)
{
boost::mutex::scoped_lock scoped_lock(other.m_mutex);
m_value = other.m_value;
}
// For assignment we need to synchronize both objects!
const counter&amp; operator=(const counter&amp; other)
{
if (this == &amp;other)
return *this;
boost::mutex::scoped_lock lock1(&amp;m_mutex &lt; &amp;other.m_mutex ? m_mutex : other.m_mutex);
boost::mutex::scoped_lock lock2(&amp;m_mutex &gt; &amp;other.m_mutex ? m_mutex : other.m_mutex);
m_value = other.m_value;
return *this;
}
int value() const
{
boost::mutex::scoped_lock scoped_lock(m_mutex);
return m_value;
}
int increment()
{
boost::mutex::scoped_lock scoped_lock(m_mutex);
return ++m_value;
}
private:
mutable boost::mutex m_mutex;
int m_value;
};
</pre>
<h2>6. How can you lock a <a href="mutex_concept.html"> mutex</a> member in a const member function, in order to
implement the Monitor Pattern?</h2>
<p>The Monitor Pattern mutex <a href="bibliography.html#Schmidt-00">[Schmidt
00]</a> should simply be declared as mutable. See the example code above. The internal state of mutex
types could have been made mutable, with all lock calls made via const
functions, but
this does a poor job of documenting the actual semantics. Declaring a mutex member
as mutable clearly documentations the intended semantics.</p>
<h2>7. Why supply <a href="condition.html">condition variables</a> rather than <a href="rationale.html#Events">
event variables</a>?</h2>
<p>Condition variables result in user code much less prone to <a href="definitions.html#Race condition">race
conditions</a> than event variables. See <a href="rationale.html#Events">Rationale</a>
for analysis. Also see <a href="bibliography.html#Hoare-74">[Hoare74]</a>
and <a href="bibliography.html#Schmidt-00">[Schmidt
00]</a>.</p>
<h2>8. Why isn't thread cancellation or termination provided?</h2>
<p>There's a valid need for thread termination, so at some point Boost.Threads
probably will include it, but only after we can find a truly safe (and portable)
mechanism for this concept.</p>
<h2>9. Is it safe for threads to share automatic storage duration (stack)
objects via pointers or references?</h2>
<p>Only if you can guarantee that the lifetime of the stack object will not end
while other threads might still access the object. Thus the safest practice is
to avoid sharing stack objects, particularly in designs where threads are
created and destroyed dynamically. Restrict sharing of stack objects to simple
designs with very clear and unchanging function and thread lifetimes. (Suggested
by Darryl Green).</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

87
doc/index.html Normal file
View File

@@ -0,0 +1,87 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<title>Boost.Threads, Index</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Documentation Map</h2>
</td>
</tr>
</table>
<hr>
<h2>Contents</h2>
<ul>
<li><a href="overview.html">Overview</a></li>
<li>Class <a href="semaphore.html">semaphore</a></li>
<li><a href="mutex_concept.html">Mutex Concepts</a></li>
<ul>
<li><a href="mutex_concept.html#Mutex">Mutex</a></li>
<li><a href="mutex_concept.html#TryMutex">TryMutex</a></li>
<li><a href="mutex_concept.html#TimedMutex">TimedMutex</a></li>
</ul>
<li>Mutex Classes</li>
<ul>
<li><a href="mutex.html">mutex / try_mutex / timed_mutex</a></li>
<li><a href="recursive_mutex.html">recursive_mutex / recursive_try_mutex / recursive_timed_mutex</a></li>
</ul>
<li><a href="lock_concept.html">Lock Concepts</a></li>
<ul>
<li><a href="lock_concept.html#Lock">Lock</a></li>
<li><a href="lock_concept.html#ScopedLock">ScopedLock</a></li>
<li><a href="lock_concept.html#ScopedTryLock">ScopedTryLock</a></li>
<li><a href="lock_concept.html#ScopedTimedLock">ScopedTimedLock</a></li>
</ul>
<li>Lock Classes</li>
<ul>
<li><a href="scoped_lock.html">scoped_lock</a></li>
<li><a href="scoped_try_lock.html">scoped_try_lock</a></li>
<li><a href="scoped_timed_lock.html">scoped_timed_lock</a></li>
</ul>
<li>Class <a href="condition.html">condition</a></li>
<li>Class <a href="thread_specific_ptr.html">thread_specific_ptr</a></li>
<li>Class <a href="thread.html">thread</a></li>
<li>Class <a href="thread_group.html">thread_group</a></li>
<li>Class <a href="xtime.html">xtime</a></li>
<li>Class <a href="lock_error.html">lock_error</a></li>
<li>Class <a href="thread_resource_error.html">thread_resource_error</a></li>
<li>Routine <a href="call_once.html">call_once</a></li>
<li><a href="config.html">Configuration Information</a></li>
<li><a href="introduction.html">Introduction to design</a></li>
<li><a href="rationale.html">Rationale for design decisions</a></li>
<li><a href="definitions.html">Definitions</a></li>
<li><a href="faq.html">Frequently Asked Questions</a></li>
<li><a href="bibliography.html">Bibliography</a></li>
<li><a href="acknowledgements.html">Acknowledgements</a></li>
</ul>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p>©<i> Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001</i></p>
<p>Permission to use, copy, modify, distribute and sell this software
and its documentation for any purpose is hereby granted without fee,
provided that the above copyright notice appear in all copies and
that both that copyright notice and this permission notice appear
in supporting documentation. William E. Kempf makes no representations
about the suitability of this software for any purpose.
It is provided &quot;as is&quot; without express or implied warranty.</p>
</body>
</html>

151
doc/introduction.html Normal file
View File

@@ -0,0 +1,151 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<title>Boost.Threads, Introduction</title>
</head>
<body bgcolor="#ffffff" link="#0000ff" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><IMG height=86 alt="C++ Boost" src="../../../c++boost.gif" width=277></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Introduction</h2>
</td>
</tr>
</table>
<hr>
<h3>Motivation</h3>
<p>With client/server and three-tier architectures becoming common place in today's
world, it's becoming increasingly important for programs to be able to handle parallel
processing. Modern day operating systems usually provide some support for this
through native thread APIs. Unfortunately, writing portable code that makes use
of parallel processing in C++ is made very difficult by a lack of a standard interface
for these native APIs. Further, these APIs are almost universally C APIs and fail to
take advantage of C++'s strengths, or to address C++'s issues.</p>
<p>The <b>Boost.Threads</b> library is an attempt to define a portable interface for writing
parallel processes in C++.</p>
<h3>Goals</h3>
<p>The <b>Boost.Threads</b> library has several goals that should help to set it apart from
other solutions. These goals are listed in order of precedence with full descriptions
below.<p>
<ul>
<li><b>Portability</b>
<p><b>Boost.Threads</b> was designed to be highly portable. The goal is for the
interface to be easily implemented on any platform that supports threads,
and possibly even on platforms without native thread support.</p>
<li><b>Safety</b>
<p><b>Boost.Threads</b> was designed to be as safe as possible. Writing
<a href="definitions.html#Thread-safe">thread-safe</a>
code is very difficult and successful libraries must strive to insulate
the programmer from dangerous constructs as much as possible. This is accomplished
in several ways:</p>
<ul>
<li><p align="left">C++ language features are used make correct usage easy (if possible,
the default) and error-prone impossible or at least more difficult.
For example, see the <a href="mutex_concept.html">Mutex</a> and <a href="lock_concept.html">Lock</a>
designs, and how note how they interact.</p></li>
<li>
<p align="left">Certain traditional concurrent programming features
are considered so error-prone that they are not provided at all. For
example, see the <a href="rationale.html#Events">Events Not Provided</a>
rationale.</p>
</li>
<li>
<p align="left">Dangerous features, or features which may be misused,
are identified as such in the documentation to make users aware of
potential pitfalls. For example, see <a href="semaphore.html#Danger">Semaphore</a>.</p>
</li>
</ul>
<li><b>Flexibility</b>
<p><b>Boost.Threads</b> was designed to be flexible. This goal is often at odds
with <i>safety</i>. When functionality might be compromised by the desire
to keep the interface safe, <b>Boost.Threads</b> has been designed to provide
the functionality, but to make it's use prohibitive for general use.</p>
<li><b>Efficiency</b>
<p><b>Boost.Threads</b> was designed to be as efficient as possible. When building
a library on top of another library there is always a danger that the result
will be so much slower than the "native" API that programmers are inclined
to ignore the higher level API. <b>Boost.Threads</b> was designed to minimize the
chances of this occurring. The interfaces have been crafted to allow an
implementation the greatest chance of being as efficient as possible. This
goal is often at odds with the goal for <i>safety</i>. Every effort was made to
ensure efficient implementations, but when in conflict <i>safety</i> has always taken
precedence.</p>
</li>
</ul>
<h3>Iterative Phases</h3>
<p>Another goal of <b>Boost.Threads</b> was to take a dynamic, iterative
approach in its development. The computing industry is still exploring the concepts of parallel programming.
Most thread libraries supply only simple primitive concepts for thread synchronization.
These concepts are very simple, but they are very difficult to use safely or to provide
formal proofs for constructs built on top of them. Until recently, these primitives
were "state of the art" and the only concepts available to programmers. Recently
there has been a lot of research in other concepts, such as in "Communicating Sequential
Processes." <b>Boost.Threads</b> was designed in iterative steps, providing the building
blocks necessary for the next step, and giving the researcher the tools necessary to
explore new concepts in a portable manner.</p>
<p>Given the goal of following a dynamic, iterative approach <b>Boost.Threads</b> shall go through
several growth cycles. Each phase in its development shall be roughly documented here.</p>
<h4>Phase 1, Synchronization Primitives</h4>
<p>Boost is all about providing high quality libraries with implementations for many platforms.
Unfortunately, there's a big problem faced by developers wishing to supply such high quality
libraries, namely thread-safety. The C++ standard doesn't address threads at all, but real
world programs often make use of native threading support. A portable library that doesn't
address the issue of thread-safety is there for not much help to a programmer who wants to
use the library in his multi-threaded application. So there's a very great need for portable
primitives that will allow the library developer to create <a href="definitions.html#Thread-safe">thread-safe</a>
implementations. This
need far out weighs the need for portable methods to create and manage threads.</p>
<p>Because of this need, the first phase of <b>Boost.Threads</b> focuses solely on providing
portable primitive concepts for thread synchronization. Types provided in this phase include
the <A href="semaphore.html">semaphore</a>, <A href="mutex.html">mutex/try_mutex/timed_mutex</a>,
<A href="recursive_mutex.html">recursive_mutex/recursive_try_mutex/recursive_timed_mutex</a>,
<A href="scoped_lock.html">scoped_lock</a>, <A href="scoped_try_lock.html">scoped_try_lock</a>,
<A href="scoped_timed_lock.html">scoped_timed_lock</a> and <A href="lock_error.html">lock_error</a>.
These are considered the "core" synchronization primitives, though there are others that will
be added in later phases.</p>
<h4>Phase 2, Thread Management and Thread Specific Storage</h4>
<p>This phase addresses the creation and management of threads and provides a mechanism for
thread specific storage (data associated with a thread instance). Thread management is a tricky
issue in C++, so this phase addresses only the basic needs of multi-threaded program. Later
phases are likely to add additional functionality in this area. This phase of <b>Boost.Threads</b>
adds the <A href="thread.html">thread</a> and
<A href="thread_specific_ptr.html">thread_specific_ptr</a> types. With these additions
the <b>Boost.Threads</b> library can be considered minimal but complete.</p>
<h4>The Next Phase</h4>
<p>The next phase will address more advanced synchronization concepts, such as read/write mutexes
and barriers.</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <A href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

242
doc/lock_concept.html Normal file
View File

@@ -0,0 +1,242 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<title>Boost.Threads, Lock Concept</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Lock Concepts</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Requirements">Concept Requirements</a><br>
<a href="#Lock">Lock Concept</a><br>
<a href="#ScopedLock">ScopedLock Concept</a><br>
<a href="#ScopedTryLock">ScopedTryLock Concept</a><br>
<a href="#ScopedTimedLock">ScopedTimedLock Concept</a><br>
<a href="#Models">Models</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The lock concepts provide exception safe means for locking and unlocking a
<a href="mutex_concept.html">mutex model</a>. In other words they are an
implementation of the <i>Scoped Locking</i>
<a href="bibliography.html#Schmidt 00">[Schmidt 00]</a> pattern. The
<a href="#ScopedLock">ScopedLock</a> concept, with
<a href="#ScopedTryLock">ScopedTryLock</a> and
<a href="#ScopedTimedLock">ScopedTimedLock</a> refinements, formalize the
requirements.</p>
<p>Lock models are constructed with a reference to a
<a href="mutex_concept.html">mutex model</a> and typically acquire ownership of the
<a href="mutex_concept.html">mutex model</a> by setting its state to locked. They also
ensure ownership is relinquished in the destructor. Lock models also expose functions
to query the lock status and to manually lock and unlock the
<a href="mutex_concept.html">mutex model</a>.</p>
<p>Instances of lock models are meant to be short lived, expected to be used at block
scope only. The lock models are not
<a href="definitions.html#Thread-safe">thread-safe</a>. Lock models must maintain state
to indicate whether or not they've been locked and this state is not protected by any
synchronization concepts. For this reason an instance of a lock model should never be
shared between multiple threads.</p>
<h2>Concept <a name="Requirements">Requirements</a></h2>
<p>[For documentation purposes, portions of the concept requirements are
repeated in the documentation for specific lock classes. Those copies need
to be kept in sync with the requirements here.]</p>
<h3><a name="Lock">Lock</a> Concept</h3>
<p>For a <a href="#ScopedLock"> ScopedLock</a>,
<a href="#ScopedTryLock">ScopedTryLock</a>, or
<a href="#ScopedTimedLock">ScopedTimedLock</a> type <code>L</code> and an object
<code>lk</code> and const object <code>clk</code> of that type, the following expressions
must be well-formed and have the indicated effects.</p>
<p>The Lock concept is used as a base for the <a href="#ScopedLock">ScopedLock</a>,
<a href="#ScopedTryLock">ScopedTryLock</a>, and
<a href="#ScopedTimedLock">ScopedTimedLock</a> refinements. The associated mutex type
is as specified for each of those refinements respectively.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Expression</b></td>
<td><b>Effects</b></td>
</tr>
<tr>
<td valign="top"><code>(&amp;lk)-&gt;~L();</code></td>
<td><code>if (locked()) unlock();</code></td>
</tr>
<tr>
<td valign="top"><code>(&amp;clk)-&gt;operator const void*()</code></td>
<td>Returns type void*, non-zero if if the associated mutex has been locked
by <code>clk</code>, otherwise 0.</td>
</tr>
<tr>
<td valign="top"><code>clk.locked()</code></td>
<td>Returns a <code>bool</code>, <code>(&amp;clk)-&gt;operator const void*() != 0</code></td>
</tr>
<tr>
<td valign="top"><code>lk.lock()</code></td>
<td>Throws lock_error if locked(). If the associated mutex is already locked by some other
thread, places the current thread in the
<a href="definitions.html#State">Blocked</a> state until the associated mutex is
unlocked, after which the current thread is placed in the
<a href="definitions.html#State">Ready</a> state, eventually to be returned to the
<a href="definitions.html#State">Running</a> state.<br>
Postcondition: locked()</td>
</tr>
<tr>
<td valign="top"><code>lk.unlock()</code></td>
<td>If !locked(), throws lock_error, otherwise unlocks the associated mutex.<br>
Postcondition: !locked()</td>
</tr>
</table>
<h3><a name="ScopedLock">ScopedLock</a> Concept</h3>
<p>A ScopedLock must meet the <a href="#Lock">Lock</a> requirements. For a ScopedLock
type <code>L</code> and an object <code>lk</code> of that type,
and an object <code>m</code> of a type meeting the
<a href="mutex_concept.html#Mutex">Mutex</a> requirements, and an object <code>b</code>
of type <code>bool</code>, the following expressions must be well-formed and have the
indicated effects.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Expression</b></td>
<td><b>Effects</b></td>
</tr>
<tr>
<td valign="top"><code>L lk(m);</code></td>
<td>Constructs an object <code>lk</code>, and associates mutex <code>m</code> with
it, then calls <code>lock()</code></td>
</tr>
<tr>
<td valign="top"><code>L lk(m,b);</code></td>
<td>Constructs an object <code>lk</code>, and associates mutex <code>m</code> with
it, then if <code>b</code>, calls <code>lock()</code></td>
</tr>
</table>
<h3><a name="ScopedTryLock">ScopedTryLock</a> Concept</h3>
<p>A ScopedTryLock must meet the <a href="#Lock">Lock</a> requirements. For a
ScopedTryLock type <code>L</code> and an object <code>lk</code> of that type,
and an object <code>m</code> of a type meeting the
<a href="mutex_concept.html#TryMutex">TryMutex</a> requirements, and an object
<code>b</code> of type <code>bool</code>, the following expressions must be well-formed
and have the indicated effects.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Expression</b></td>
<td><b>Effects</b></td>
</tr>
<tr>
<td valign="top"><code>L lk(m);</code></td>
<td>Constructs an object <code>lk</code>, and associates mutex <code>m</code> with
it, then calls <code>try_lock()</code></td>
</tr>
<tr>
<td valign="top"><code>L lk(m,b);</code></td>
<td>Constructs an object <code>lk</code>, and associates mutex <code>m</code> with
it, then if <code>b</code>, calls <code>lock()</code></td>
</tr>
<tr>
<td valign="top"><code>lk.try_lock()</code></td>
<td>If locked(), throws <code>lock_error</code>. Makes a non-blocking attempt to
lock the associated mutex, returning <code>true</code> if the lock attempt is
successful, otherwise <code>false</code>.</td>
</tr>
</table>
<h3><a name="ScopedTimedLock">ScopedTimedLock</a> Concept</h3>
<p>A ScopedTimedLock must meet the <a href="#Lock">Lock</a> requirements. For a
ScopedTimedLock type <code>L</code> and an object <code>lk</code> of that type,
and an object <code>m</code> of a type meeting the
<a href="mutex_concept.html#TimedMutex">TimedMutex</a> requirements, and an object
<code>b</code> of type <code>bool</code>, and an object <code>t</code> of type
<code><a href="xtime.html">xtime</a></code>, the following expressions must be well-formed and have the indicated
effects.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Expression</b></td>
<td><b>Effects</b></td>
</tr>
<tr>
<td valign="top"><code>L lk(m,t);</code></td>
<td>Constructs an object <code>lk</code>, and associates mutex <code>m</code> with
it, then calls <code>timed_lock(t)</code></td>
</tr>
<tr>
<td valign="top"><code>L lk(m,b);</code></td>
<td>Constructs an object <code>lk</code>, and associates mutex <code>m</code> with
it, then if <code>b</code>, calls <code>lock()</code></td>
</tr>
<tr>
<td valign="top"><code>lk.timed_lock(t)</code></td>
<td>If locked(), throws lock_error. Makes a blocking attempt to lock the
associated mutex, and returns <code>true</code> if successful within the specified
time <code>t</code>, otherwise <code>false</code>.</td>
</tr>
</table>
<h2><a name="Models">Models</a></h2>
<p><b>Boost.Threads</b> currently supplies three classes which model lock concepts.</p>
<p>These classes are normally accessed via typedefs of the same name supplied by
a <a href="mutex_concept.html">mutex model</a>.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Concept</b></td>
<td><b>Refines</b></td>
<td><b>Classes Modeling the Concept</b></td>
</tr>
<tr>
<td><a href="#ScopedLock">ScopedLock</a></td>
<td>&nbsp;</td>
<td><a href="scoped_lock.html">scoped_lock</a></td>
</tr>
<tr>
<td><a href="#ScopedTryLock">ScopedTryLock</a></td>
<td><a href="#ScopedLock">ScopedLock</a></td>
<td><a href="scoped_try_lock.html">scoped_try_lock</a> </td>
</tr>
<tr>
<td><a href="#ScopedTimedLock">ScopedTimedLock</a></td>
<td><a href="#ScopedLock">ScopedLock</a></td>
<td><a href="scoped_timed_lock.html">scoped_timed_lock</a></td>
</tr>
</table>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

111
doc/lock_error.html Normal file
View File

@@ -0,0 +1,111 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, lock_error</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">lock_error</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#Synopsis">Synopsis</a><br>
<a href="#Members">Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <tt>lock_error</tt> class defines an exception type thrown to indicate a
locking related error has been detected. Examples of such errors include a lock
operation which can be determined to result in a deadlock, or unlock operations
attempted by a thread that does not own the lock.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost
class lock_error : public std::runtime_error
{
public:
lock_error();
};
}
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
lock_error();
</pre>
<p>Constructs a <tt>lock_error</tt> object.</p>
<hr>
<h2><a name="Example">Example</a> Usage</h2>
<pre>
#include <a href="../../../boost/thread/mutex.hpp">&lt;boost/thread/mutex.hpp&gt;</a>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
#include &lt;iostream&gt;
int main(int, char*[])
{
boost::mutex mutex;
boost::mutex::scoped_lock scoped_lock(mutex);
try
{
boost::mutex::scoped_lock deadlock(mutex);
std::cout &lt;&lt; &quot;lock succeeded&quot; &lt;&lt; std::endl;
}
catch (boost::lock_error&amp; err)
{
std::cout &lt;&lt; err.what() &lt;&lt; &quot; - deadlock occurred.&quot; &lt;&lt; std::endl;
}
return 0;
}
</pre>
<p>The output is:</p>
<pre>
thread lock error - deadlock occurred.
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

319
doc/mutex.html Normal file
View File

@@ -0,0 +1,319 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, mutex</title>
</head>
<body bgcolor="#ffffff" link="#0000ff" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><IMG alt="C++ Boost" src="../../../c++boost.gif" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">mutex<br>
try_mutex<br>
timed_mutex</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#mutex Synopsis">Class mutex Synopsis</a><br>
<a href="#mutex Members">Class mutex Members</a><br>
<a href="#try_mutex Synopsis">Class try_mutex Synopsis</a><br>
<a href="#try_mutex Members">Class try_mutex Members</a><br>
<a href="#timed_mutex Synopsis">Class timed_mutex Synopsis</a><br>
<a href="#timed_mutex Members">Class timed_mutex Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <tt><a href="#mutex Synopsis">mutex</a></tt>, <tt><a href="#try_mutex Synopsis">try_mutex</a></tt> and <tt><a href="#timed_mutex Synopsis">timed_mutex</a></tt> classes define full featured
models of the <a href="mutex_concept.html#Mutex">Mutex</a>, <a href="mutex_concept.html#TryMutex">TryMutex</a>,
and <a href="mutex_concept.html#TimedMutex">TimedMutex</a> concepts. These types should be used to
non-recursively synchronize access to
shared resources. For recursive locking mechanics, see <a href="recursive_mutex.html">recursive
mutexes</a>.</p>
<p>Each class supplies one or more typedefs for lock types which model matching
lock concepts. For the best possible performance you should use the mutex class that supports
the minimum set of lock
types that you need.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Mutex Class</b></td>
<td><b>Lock name</b></td>
<td><b>Implementation defined Lock Type</b></td>
<td><b>&nbsp;Lock Concept</b></td>
</tr>
<tr>
<td valign="top"><a href="#mutex Synopsis"><code>mutex</code></a></td>
<td valign="middle"><code>scoped_lock</code></td>
<td valign="middle"><code><a href="scoped_lock.html">boost::</a></code><a href="scoped_lock.html"><code>detail::thread::scoped_lock&lt;mutex&gt;</code></a></td>
<td valign="middle"><a href="lock_concept.html#ScopedLock">ScopedLock</a></td>
</tr>
<tr>
<td valign="top"><tt><a href="#try_mutex Synopsis">try_mutex</a></tt> </td>
<td valign="middle"><code>scoped_lock<br>
scoped_try_lock</code></td>
<td valign="middle"><code><a href="scoped_lock.html">boost::</a></code><a href="scoped_lock.html"><code>detail::thread::scoped_lock&lt;try_mutex&gt;<br>
</code></a><code><a href="scoped_try_lock.html">boost::detail::thread::scoped_try_lock&lt;try_mutex&gt;</a></code></td>
<td valign="middle"><a href="lock_concept.html#ScopedLock">ScopedLock</a><br>
<a href="lock_concept.html#ScopedTryLock">ScopedTryLock</a></td>
</tr>
<tr>
<td valign="top"><code><a href="#timed_mutex Synopsis">timed_mutex</a></code> </td>
<td valign="middle"><code>scoped_lock<br>
scoped_try_lock<br>
scoped_timed_lock</code></td>
<td valign="middle"><code><a href="scoped_lock.html">boost::</a></code><a href="scoped_lock.html"><code>detail::thread::scoped_lock&lt;timed_mutex&gt;</code></a><br>
<code><a href="scoped_try_lock.html">boost::</a></code><a href="scoped_try_lock.html"><code>detail::thread::scoped_try_lock&lt;timed_mutex&gt;</code></a><br>
<code><a href="scoped_timed_lock.html">boost::</a></code><a href="scoped_timed_lock.html"><code>detail::thread::scoped_timed_lock&lt;timed_mutex&gt;</code></a></td>
<td valign="middle"><a href="lock_concept.html#ScopedLock">ScopedLock</a><br>
<a href="lock_concept.html#ScopedTryLock">ScopedTryLock</a><br>
<a href="lock_concept.html#ScopedTimedLock">ScopedTimedLock</a></td>
</tr>
</table>
<p>The <tt>mutex</tt>, <tt>try_mutex</tt> and <tt>timed_mutex</tt> classes use an <tt>Unspecified</tt>
<A href="mutex_concept.html#LockingStrategies">locking strategy</a>, so attempts to recursively lock
them or attempts to unlock them by threads that don't own a lock on them result in <b>undefined behavior</b>.
This strategy allows implementations to be as efficient as possible on any given platform. It is, however,
recommended that implementations include debugging support to detect misuse when <tt>NDEBUG</tt> is
not defined.</p>
<p>Like all the <b>Boost.Threads</b> <A href="mutex_concept.html">mutex models</a>, the <tt>mutex</tt>,
<tt>try_mutex</tt> and <tt>timed_mutex</tt> leave the
<A href="mutex_concept.html#SchedulingPolicies">scheduling policy</a> as <tt>Unspecified</tt>.
Programmers should assume that threads waiting for a lock on objects of these types
acquire
the lock in a random order, even though the specific behavior for a given platform may be different.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/mutex.hpp">&lt;boost/thread/mutex.hpp&gt;</a>
</pre>
<h2>Class <a name="mutex Synopsis"> mutex Synopsis</a></h2>
<pre>
namespace boost
{
class mutex : private <a href="../../utility/utility.htm">boost::noncopyable</a> // Exposition only.
// Class mutex meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_lock;
mutex();
~mutex();
};
}
</pre>
<h2>
Class <a name="mutex Members">mutex Members</a>
</h2>
<hr>
<h3>Constructor</h3>
<pre>
mutex();
</pre>
<p><b>Postconditions: </b><code>*this</code> is in the unlocked state.</p>
<hr>
<h3>Destructor</h3>
<pre>
~mutex();
</pre>
<p><b>Requires:</b> <code>*this</code> is in the unlocked state.</p>
<p><b>Effects:</b> Destroys <code>*this</code>.</p>
<p><b>Dangers:</b> Destruction of a locked mutex is a serious programming error
resulting in undefined behavior such as a program crash.</p>
<hr>
<h2>
Class <a name="try_mutex Synopsis">try_mutex Synopsis</a>
</h2>
<pre>
namespace boost
{
class try_mutex : private boost::noncopyable // Exposition only.
// Class try_mutex meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_lock;
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_try_lock;
try_mutex();
~try_mutex();
};
}
</pre>
<h2>Class <a name="try_mutex Members">try_mutex Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
try_mutex();
</pre>
<p><b>Postconditions: </b><code>*this</code> is in the unlocked state.</p>
<hr>
<h3>Destructor</h3>
<pre>
~try_mutex();
</pre>
<p><b>Requires:</b> <code>*this</code> is in the unlocked state.</p>
<p><b>Effects:</b> Destroys <code>*this</code>.</p>
<p><b>Dangers:</b> Destruction of a locked mutex is a serious programming error
resulting in undefined behavior such as a program crash.</p>
<hr>
<h2>
Class <a name="timed_mutex Synopsis">timed_mutex Synopsis</a>
</h2>
<pre>
namespace boost
{
class timed_mutex : private boost::noncopyable // Exposition only.
// Class timed_mutex meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_lock;
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_try_lock;
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_timed_lock;
timed_mutex();
~timed_mutex();
};
}
</pre>
<h2>Class <a name="timed_mutex Members">timed_mutex Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
timed_mutex();
</pre>
<p><b>Postconditions: </b><code>*this</code> is in the unlocked state.</p>
<hr>
<h3>Destructor</h3>
<pre>
~timed_mutex();
</pre>
<p><b>Requires:</b> <code>*this</code> is in the unlocked state.</p>
<p><b>Effects:</b> Destroys <code>*this</code>.</p>
<p><b>Dangers:</b> Destruction of a locked mutex is a serious programming error
resulting in undefined behavior such as a program crash.</p>
<hr>
<h2><a name="Example">Example</a> Usage</h2>
<pre>
#include <a href="../../../boost/thread/mutex.hpp">&lt;boost/thread/mutex.hpp&gt;</a>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
#include &lt;iostream&gt;
boost::mutex io_mutex; // The iostreams are not guaranteed to be <a href="definitions.html#Thread-safe">thread-safe</a>!
class counter
{
public:
counter() : count(0) { }
int increment() {
boost::mutex::scoped_lock scoped_lock(mutex);
return ++count;
}
private:
boost::mutex mutex;
int count;
};
counter c;
void change_count(void*)
{
int i = c.increment();
boost::mutex::scoped_lock scoped_lock(io_mutex);
std::cout &lt;&lt; &quot;count == &quot; &lt;&lt; i &lt;&lt; std::endl;
}
int main(int, char*[])
{
const int num_threads = 4;
boost::thread_group thrds;
for (int i=0; i &lt; num_threads; ++i)
thrds.create_thread(&amp;change_count, 0);
thrds.join_all();
return 0;
}
</pre>
<p>The output is:</p>
<pre>
count == 1
count == 2
count == 3
count == 4
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <A href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

275
doc/mutex_concept.html Normal file
View File

@@ -0,0 +1,275 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<title>Boost.Threads, Mutex Concept</title>
</head>
<body bgcolor="#ffffff" link="#0000ff" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><IMG height=86 alt="C++ Boost" src="../../../c++boost.gif" width=277></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Mutex Concepts</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#LockingStrategies">Locking Strategies</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#Recursive">Recursive</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#CheckedStrategy">Checked</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#UncheckedStrategy">Unchecked</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#UnspecifiedStrategy">Unspecified</a><br>
<a href="#SchedulingPolicies">Scheduling Policies</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#FIFO">FIFO</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#Priority Driven">Priority Driven</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#UndefinedScheduling">Undefined</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#UnspecifiedScheduling">Unspecified</a><br>
<a href="#Requirements">Concept Requirements</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#Mutex">Mutex Concept</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#TryMutex">TryMutex Concept</a><br>
&nbsp;&nbsp;&nbsp;&nbsp;<a href="#TimedMutex">TimedMutex Concept</a><br>
<a href="#Models">Models</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>A mutex (short for mutual-exclusion) concept serializes access to
a resource shared between multiple threads. The <a href="#Mutex">Mutex</a>
concept, with <a href="#TryMutex">TryMutex</a> and <a href="#TimedMutex">TimedMutex</a>
refinements, formalize the requirements. A model that implements Mutex and its
refinements has two states: <b> locked</b> and <b>unlocked</b>. Before using a
shared resource, a thread locks a Boost.Threads mutex model object,
insuring <a href="definitions.html#Thread-safe">thread-safe</a> access to the shared
resource. When use of the shared resource is complete, the thread unlocks the mutex
model object, allowing another thread to acquire the lock and use the shared resource.</p>
<p>Traditional C thread APIs, like Pthreads or the Windows thread APIs, expose
functions to lock and unlock a mutex model. This is dangerous since it's easy to forget
to unlock a locked mutex. When the flow of control is complex, with multiple return
points, the likelihood of forgetting to unlock a mutex model would become even greater.
When exceptions are thrown, it becomes nearly impossible to ensure that the mutex is
unlocked properly when using these traditional API's. The result is
<a href="definitions.html#Deadlock">deadlock</a>.</p>
<p>Many C++ threading libraries use a pattern known as <i>Scoped Locking</i>
<a href="bibliography.html#Schmidt 00">[Schmidt 00]</a> to free the programmer from the
need to explicitly lock and unlock mutexes. With this pattern, a
<A href="lock_concept.html">lock concept</A> is employed where the lock model's
constructor locks the associated mutex model and the destructor automatically does the
unlocking. The <b>Boost.Threads</b> library takes this pattern to the extreme in that
lock concepts are the only way to lock and unlock a mutex model: lock and unlock
functions are not exposed by any <b>Boost.Threads </b>mutex models. This helps to
ensure safe usage patterns, especially when code throws exceptions.</p>
<h2><a name="LockingStrategies">Locking Strategies</a></h2>
<p>Every mutex model follows one of several locking strategies. These strategies
define the semantics for the locking operation when the calling thread already
owns a lock on the mutex model.</p>
<h3><a name="Recursive">Recursive</a></h3>
<p>With a recursive locking strategy when a thread attempts to acquire a lock on
the mutex model for which it already owns a lock, the operation is successful.
Note the distinction between a thread, which may have multiple locks outstanding
on a recursive mutex, and a lock object, which even for a recursive mutex cannot
have its lock() function called multiple times without first calling unlock().</p>
<p>Internally a lock count is maintained and the owning thread must unlock the
mutex model the same number of times that it's locked it before the mutex model's
state returns to unlocked. Since mutex models in <b>Boost.Threads</b> expose
locking functionality only through lock concepts, a thread will always unlock a mutex
model the same number of times that it locked it. This helps to eliminate a whole set
of errors typically found in traditional C style thread APIs.</p>
<p>Classes <A href="recursive_mutex.html">recursive_mutex</A>,
<A href="recursive_mutex.html">recursive_try_mutex</A> and
<A href="recursive_mutex.html">recursive_timed_mutex</A> use this locking strategy.</p>
<h3><a name="CheckedStrategy">Checked</a></h3>
<p>With a checked locking strategy when a thread attempts to acquire a lock on
the mutex model for which the thread already owns a lock, the operation will fail with
some sort of error indication. Further, attempts by a thread to unlock a mutex
that was not locked by the thread will also return some sort of error indication.
In <b>Boost.Threads</b>, an exception of type <A href="lock_error.html">lock_error</A>
would be thrown in these cases.</p>
<p><b>Boost.Threads</b> does not currently provide any mutex models that use this
strategy.</p>
<h3><a name="UncheckedStrategy">Unchecked</a></h3>
<p>With an unchecked locking strategy when a thread attempts to acquire a lock
on the mutex model for which the thread already owns a lock the operation will
<a href="definitions.html#Deadlock">deadlock</a>. In general this locking strategy is
less safe than a checked or recursive strategy, but it's also a faster strategy and so
is employed by many libraries.</p>
<p><b>Boost.Threads</b> does not currently provide any mutex models that use this
strategy.</p>
<h3><a name="UnspecifiedStrategy">Unspecified</a></h3>
<p>With an unspecified locking strategy, when a thread attempts to acquire a lock
on a mutex model for which the thread already owns a lock the operation results in
<b>undefined behavior</b>. When a mutex model has an unspecified locking strategy the
programmer must assume that the mutex model instead uses an unchecked strategy.</p>
<p>In general a mutex model with an unspecified locking strategy is unsafe, and it
requires programmer discipline to use the mutex model properly. However, this strategy
allows an implementation to be as fast as possible with no restrictions on its
implementation. This is especially true for portable implementations that wrap the
native threading support of a platform. For this reason, the classes
<A href="mutex.html">mutex</A>, <A href="mutex.html">try_mutex</A> and
<A href="mutex.html">timed_mutex</A> use this locking strategy despite the lack of
safety.</p>
<h2><a name="SchedulingPolicies">Scheduling Policies</a></h2>
<p>Every mutex model follows one of several scheduling policies. These policies
define the semantics when the mutex model is unlocked and there is more than one
thread waiting to acquire a lock. In other words, the policy defines which waiting
thread shall acquire the lock.</p>
<h3><a name="FIFO">FIFO</a></h3>
<p>With a FIFO scheduling policy, threads waiting for the lock will acquire it in
a first come first serve order (or First In First Out). This can help prevent a
high priority thread from starving lower priority threads that are also waiting
on the mutex lock.</p>
<h3><a name="Priority Driven">Priority Driven</a></h3>
<p>With a Priority Driven scheduling policy, the thread with the highest priority
acquires the lock. Note that this means that low-priority threads may never acquire
the lock if the mutex model has high contention and there is always at least one
high-priority thread waiting. This is known as thread starvation. When multiple threads
of the same priority are waiting on the mutex lock one of the other scheduling
priorities will determine which thread shall acquire the lock.</p>
<h3><a name="UndefinedScheduling">Undefined</a></h3>
<p>Threads acquire the lock in no particular order. Users should assume that
low-priority threads may wait indefinitely, and that threads of the same
priority acquire the lock in essentially random order.</p>
<h3><a name="UnspecifiedScheduling">Unspecified</a></h3>
<p>The mutex model does not specify which scheduling policy is used. The programmer
must assume that an undefined scheduling policy is used. In order to ensure portability,
all <b>Boost.Threads</b> mutex models use an unspecified scheduling policy.</p>
<h2>Concept <a name="Requirements">Requirements</a></h2>
<h3><a name="Mutex">Mutex</a> Concept</h3>
<p>A Mutex object has two states: locked and unlocked. Mutex object state can only be
determined by an object meeting the <a href="lock_concept.html#ScopedLock">ScopedLock</a>
requirements and constructed for the Mutex object.</p>
<p>A Mutex is <a href="../../utility/utility.htm#Class noncopyable">noncopyable</a>.</p>
<p>For a Mutex type M and an object m of that type, the following expressions must be
well-formed and have the indicated effects.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Expression</b></td>
<td><b>Effects</b></td>
</tr>
<tr>
<td><code>M m;</code></td>
<td>Constructs a mutex object m. Post-condition: m is unlocked.</td>
</tr>
<tr>
<td><code>(&amp;m)-&gt;~M();</code></td>
<td>Precondition: m is unlocked. Destroys a mutex object m.</td>
</tr>
<tr>
<td><code>M::scoped_lock</code></td>
<td>A type meeting the <a href="lock_concept.html#ScopedLock">ScopedLock</a>
requirements.</td>
</tr>
</table>
<h3><a name="TryMutex">TryMutex</a> Concept</h3>
<p>A TryMutex must meet the <a href="#Mutex"> Mutex</a> requirements. In addition, for a
TryMutex type M and an object m of that type, the following expressions must be
well-formed and have the indicated effects.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Expression</b></td>
<td><b>Effects</b></td>
</tr>
<tr>
<td><code>M::scoped_try_lock</code></td>
<td>A type meeting the <a href="lock_concept.html#ScopedTryLock">ScopedTryLock</a>
requirements.</td>
</tr>
</table>
<h3><a name="TimedMutex">TimedMutex</a> Concept</h3>
<p>A TimedMutex must meet the <a href="#TryMutex"> TryMutex</a> requirements. In addition, for a
TimedMutex type M and an object m of that type, the following
expressions must be well-formed and have the indicated effects.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Expression</b></td>
<td><b>Effects</b></td>
</tr>
<tr>
<td><code>M::scoped_timed_lock</code></td>
<td>A type meeting the <a href="lock_concept.html#ScopedTimedLock">ScopedTimedLock</a>
requirements.</td>
</tr>
</table>
<h2><a name="Models">Models</a></h2>
<p> <b>Boost.Threads</b> currently supplies six classes which model mutex
concepts.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Concept</b></td>
<td><b>Refines</b></td>
<td><b>Classes Modeling the Concept</b></td>
</tr>
<tr>
<td valign="top"><a href="#Mutex">Mutex</a></td>
<td valign="top">&nbsp;</td>
<td><A href="mutex.html">mutex</A><br>
<A href="recursive_mutex.html">recursive_mutex</A></td>
</tr>
<tr>
<td valign="top"><a href="#TryMutex">TryMutex</a></td>
<td valign="top"><a href="#Mutex">Mutex</a></td>
<td><A href="mutex.html">try_mutex<br>
</A><A href="recursive_mutex.html">recursive_try_mutex</A> </td>
</tr>
<tr>
<td valign="top"><a href="#TimedMutex">TimedMutex</a></td>
<td valign="top"><a href="#TryMutex">TryMutex</a></td>
<td><A href="mutex.html">timed_mutex<br>
</A><A href="recursive_mutex.html">recursive_timed_mutex</A></td>
</tr>
</table>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <A href="mailto:williamkempf@hotmail.com">William E. Kempf</A>
2001 all rights reserved.</i></p>
</body>
</html>

156
doc/overview.html Normal file
View File

@@ -0,0 +1,156 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
<meta name="GENERATOR" content="Microsoft FrontPage 4.0">
<meta name="ProgId" content="FrontPage.Editor.Document">
<title>Boost.Threads Overview</title>
</head>
<body>
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><IMG height=86 alt="C++ Boost" src="../../../c++boost.gif" width=277></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Overview</h2>
</td>
</tr>
</table>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Dangers">Dangers</a><br>
<a href="#Library">C++ Standard Library usage</a><br>
<a href="#Common">Common requirements</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>Boost.Threads allows C++ programs to execute as multiple, asynchronous,
independent, threads-of-execution. Each thread has its own machine state
including program instruction counter and registers. Programs which execute as
multiple threads are call multi-threaded programs to distinguish them from
traditional single-threaded programs. <a href="definitions.html">Definitions</a>
gives a more complete description of the multi-threading execution environment.</p>
<p>Multi-threading provides several advantages:</p>
<ul>
<li>Programs which would otherwise block waiting for some external event can
continue to respond if the blocking operation is placed in a separate
thread.&nbsp; Multi-threading is usually an absolute requirement for these
programs.</li>
</ul>
<ul>
<li>Well-designed multi-threaded programs may execute faster than single-threaded
programs, particularly on multi-processor hardware.&nbsp;
Note, however, that poorly-designed multi-threaded programs are often slower
that single-threaded programs.</li>
</ul>
<ul>
<li>Some program designs may be easier to formulate using a multi-threaded
approach.
After all, the real world is asynchronous!&nbsp;</li>
</ul>
<h2><a name="Dangers">Dangers</a></h2>
<p>Beyond the errors which can occur in single-threaded programs, multi-threaded
programs are subject to additional errors:</p>
<ul>
<li><a href="definitions.html#Race condition">Race conditions</a>.
<li><a href="definitions.html#Deadlock">Deadlock</a> (sometimes called
&quot;deadly embrace&quot;)
<li><a href="definitions.html#Priority failure">Priority failures</a>
(priority inversion, infinite overtaking, starvation, etc.)</li>
</ul>
<p>Every multi-threaded program must be designed carefully to avoid race
conditions and deadlock.&nbsp; These aren't rare or exotic failures - they are
virtually guaranteed to occur unless multi-threaded code is designed to avoid
them.&nbsp; Priority failures are somewhat less common, but are none-the-less
serious.</p>
<p>The <a href="introduction.html">Boost.Threads design</a> attempts to minimize
these errors, but they will still occur unless the programmer proactively
designs to avoid them.</p>
<h3>Testing and debugging considerations</h3>
<p>Multi-threaded programs are non-deterministic. In other words, the same
program with the same input data may follow different execution paths each time
it is invoked.&nbsp; That can make testing and debugging a nightmare:</p>
<ul>
<li>Failures are often not repeatable.
<li>Probe effect causes debuggers to produce very different results from
non-debug uses.
<li>Debuggers require special support to show thread state.
<li>Tests on a single processor system may give no indication of serious
errors which would appear on multiprocessor systems, and visa versa. Thus test
cases should include a varying number of processors.&nbsp;</li>
<li>For programs which create a varying number of threads according to
workload, tests which don't span the full range of possibilities may miss
serious errors.</li>
</ul>
<h3>Getting a head start</h3>
<p>Although it might appear that multi-threaded programs are inherently
unreliable, many reliable multi-threaded programs do exist. Multi-threading
techniques are known which lead to reliable programs.</p>
<p>Design patterns for reliable multi-threaded programs, including the important
<i>monitor</i> pattern, are presented in <cite>Pattern-Oriented Software Architecture Volume 2 - Patterns for
Concurrent and Networked Objects</cite> [<a href="bibliography.html#Schmidt-00">Schmidt
00</a>].&nbsp; Many important multi-threading programming considerations
(independent of threading library) are discussed in <cite>Programming with
POSIX Threads</cite> [<a href="bibliography.html#Butenhof-97">Butenhof 97</a>].</p>
<p>Reading and study first yields a head start toward designing reliable
multi-threaded programs.</p>
<h2><a name="Library">C++ Standard Library usage in multi-threaded programs</a></h2>
<h3>Runtime libraries</h3>
<p><b>Warning:</b> Multi-threaded programs such as those using <b>Boost.Threads</b> must link to
<a href="definitions.html#Thread-safe">thread-safe</a> versions of all runtime
libraries used by the program, including the runtime library for the C++
Standard Library. Otherwise <a href="definitions.html#Race condition">race
conditions</a> will occur when multiple threads simultaneously execute runtime
library functions for <i>new</i>, <i>delete</i>, or other language features
which imply shared state.&nbsp;&nbsp;</p>
<h3>Potentially non-thread-safe functions</h3>
<p>Certain C++ Standard Library functions inherited from C are particular
problems because they hold internal state between calls:</p>
<ul>
<li>rand</li>
<li>strtok</li>
<li>asctime</li>
<li>ctime&nbsp;</li>
<li>gmtime</li>
<li>localtime</li>
</ul>
<p>It is possible to write thread-safe implementations of these by using <a href="thread_specific_ptr.html">thread-specific
storage</a>, and several C++ compiler vendors do just that.&nbsp; The technique
is well-know and is explained in [<a href="bibliography.html#Butenhof-97">Buttenhof-97</a>].</p>
<p>But at least one vendor (HP-UX) does not provide thread-safe implementations
of the above functions in their otherwise thread-safe runtime library.&nbsp;
Instead they provide replacement functions with different names and arguments.</p>
<p><b>Recommendation:</b> For the most portable, yet thread-safe code, use Boost
replacements for the problem functions.&nbsp; See the <a href="../../random/index.html">Boost
Random Number Library</a> and <a href="../../tokenizer/index.htm">Boost
Tokenizer Library</a>.</p>
<h2><a name="Common">Common</a> requirements for all Boost.Threads components</h2>
<h3>Exceptions</h3>
<p> <b>Boost.Threads</b> destructors never throw exceptions. Unless otherwise
specified, other <b>Boost.Threads</b>
functions that do not have an exception-specification may throw implementation-defined exceptions.</p>
<p>In particular, <b>Boost.Threads</b> reports failure to allocate storage by throwing an exception of type
std::bad_alloc, or a class derived from std::bad_alloc, failure to obtain
thread resources other than memory by throwing an exception of type <a href="thread_resource_error.html">boost::thread_resource_error</a>,
and certain lock related failures by throwing an exception of type <a href="lock_error.html">boost::lock_error</a></p>
<p><b>Rationale: </b>Follows the C++ Standard Library practice of allowing all
functions except destructors or other specified functions to throw exceptions on
errors.</p>
<h3><a name="NonCopyable">NonCopyable</a> requirement</h3>
<p><b>Boost.Threads</b> classes documented as meeting the NonCopyable requirement disallow copy
construction and copy assignment. For the sake of exposition, the synopsis of
such classes show private derivation from <a href="../../utility/utility.htm">boost::noncopyable</a>.
Users should not depend on this derivation, however, as implementations are free
to meet the NonCopyable requirement in other ways.</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p>© Copyright 2001 Beman Dawes</p>
</body>
</html>

431
doc/rationale.html Normal file
View File

@@ -0,0 +1,431 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<title>Boost.Threads, rationale</title>
</head>
<body bgcolor="#ffffff" link="#0000ff" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><IMG height=86 alt="C++ Boost" src="../../../c++boost.gif" width=277></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Rationale</h2>
</td>
</tr>
</table>
<hr>
<p>This page explains the rationale behind various design decisions in the <b> Boost.Threads</b>
library. Having the rationale documented here should explain how we arrived at the current
design as well as prevent future rehashing of discussions and thought processes that have
already occurred. It can also give users a lot of insight into the design process required
for this library.</p>
<h2><a name="library">Rationale for the Creation of Boost.Threads</a></h2>
<p>Processes often have a degree of "potential parallelism" and it can often be more intuitive
to design systems with this in mind. Further, these parallel processes can result in more responsive
programs. The benefits for multi-threaded programming are quite well known to most modern programmers,
yet the C++ language doesn't directly support this concept.</p>
<p>Many platforms support multi-threaded programming despite the fact that the language doesn't support
it. They do this through external libraries, which are, unfortunately, platform specific. POSIX has
tried to address this problem through the standardization of a "pthread" library. However, this
is a standard only on POSIX platforms, so its portability is limited.</p>
<p>Another problem with POSIX and other platform specific thread libraries is that they are
almost universally C based libraries. This leaves several C++ specific issues unresolved, such
as what happens when an exception is thrown in a thread. Further, there are some C++ concepts,
such as destructors, that can make usage much easier than what's available in a C library.</p>
<p>What's truly needed is C++ language support for threads. However, the C++ standards committee needs
existing practice or a good proposal as a starting point for adding this to the standard.</p>
<p>The Boost.Threads library was developed to provide a C++ developer with a portable interface
for writing multi-threaded programs on numerous platforms. There's a hope that the library can
be the basis for a more detailed proposal for the C++ standards committee to consider for inclusion
in the next C++ standard.</p>
<h2><a name="primitives">Rationale for the Low Level Primitives Supported in Boost.Threads</a></h2>
<p>The Boost.Threads library supplies a set of low level primitives for writing multi-threaded
programs, such as semaphores, mutexes and condition variables. In fact, the first release of
Boost.Threads supports only these low level primitives. However, computer
science research has shown
that use of these primitives is difficult since there's no way to mathematically prove that a
usage pattern is correct, meaning it doesn't result in race conditions or deadlocks. There
are several algebras (such as CSP, CCS and Join calculus) that have been developed to help write
provably correct parallel processes. In order to prove the correctness these processes must
be coded using higher level abstractions. So why does Boost.Threads support the lower level
concepts?</p>
<p>The reason is simple: the higher level concepts need to be implemented using at least some
of the lower level concepts. So having portable lower level concepts makes it easier to develop
the higher level concepts and will allow researchers to experiment with various techniques.</p>
<p>Beyond this theoretical application of higher level concepts, however, the fact remains that
many multi-threaded programs are written using only the lower level concepts, so they are
useful in and of themselves, even if it's hard to prove that their usage is correct. Since
many users will be familiar with these lower level concepts but be unfamiliar with any of the
higher level concepts there's also an argument for accessibility.</p>
<h2><a name="lock_objects">Rationale for the Lock Design</a></h2>
<p>Programmers who are used to multi-threaded programming issues will quickly note that the
Boost.Thread's design for mutex lock concepts is not <a href="definitions.html#Thread-safe">thread-safe</a>
(this is clearly documented
as well). At first this may seem like a serious design flaw. Why have a multi-threading primitive
that's not thread-safe itself?</p>
<p>A lock object is not a synchronization primitive. A lock object's sole responsibility is
to ensure that a mutex is both locked and unlocked in a manner that won't result in the common
error of locking a mutex and then forgetting to unlock it. This means that instances of a
lock object are only going to be created, at least in theory, within block scope and won't
be shared between threads. Only the mutex objects will be created outside of block scope and/or
shared between threads. Though it's possible to create a lock object outside of block scope and
to share it between threads to do so would not be a typical usage. Nor are there any cases when
such usage would be required.</p>
<p>Lock objects must maintain some state information. In order to allow a program to determine
if a try_lock or timed_lock was successful the lock object must retain state indicating
the success or failure of the call made in its constructor. If a lock object were to have
such state and remain thread-safe it would need to synchronize access to the state information
which would result in roughly doubling the time of most operations. Worse, since checking
the state can occur only by a call after construction we'd have a race condition if the lock
object were shared between threads.</p>
<p>So, to avoid the overhead of synchronizing access to the state information and to avoid
the race condition the Boost.Threads library simply does nothing to make lock objects thread-safe. Instead, sharing a lock object between threads results in undefined behavior. Since the
only proper usage of lock objects is within block scope this isn't a problem, and so long
as the lock object is properly used there's no danger of any multi-threading issues.</p>
<h2><a name="thread">Rationale for Non-copyable Thread Type</a></h2>
<p>Programmers who are used to C libraries for multi-threaded programming are likely to
wonder why Boost.Threads uses a non-copyable design for <a href="thread.html">boost::thread</a>. After all, the C
thread types are copyable, and you often have a need for copying them within user code.
However, careful comparison of C designs to C++ designs shows a flaw in this logic.</p>
<p>All C types are copyable. It is, in fact, not possible to make a non-copyable type in
C. For this reason types that represent system resources in C are often designed to behave
very similarly to a pointer to dynamic memory. There's an API for acquiring the resource
and an API for releasing the resources. For memory we have pointers as the type and
alloc/free for the acquisition and release APIs. For files we have FILE* as the type
and fopen/fclose for the acquisition and release APIs. You can freely copy instances of the
types but must manually manage the lifetime of the actual resource through the acquisition
and release APIs.</p>
<p>C++ designs recognize that the acquisition and release APIs are error prone and try
to eliminate possible errors by acquiring the resource in the constructor and releasing it
in the destructor. The best example of such a design is the std::iostream set of classes
which can represent the same resource as the FILE* type in C. A file is opened in the
std::fstream's constructor and closed in its destructor. However, if an iostream were
copyable it could lead to a file being closed twice, an obvious error, so the std::iostream
types are noncopyable by design. This is the same design used by boost::thread, which
is a simple and easy to understand design that's consistent with other C++ standard types.</p>
<p>During the design of boost::thread it was pointed out that it would be possible to allow
it to be a copyable type if some form of "reference management" were used, such as ref-counting
or ref-lists, and many argued for a boost::thread_ref design instead. The reasoning was
that copying "thread" objects was a typical need in the C libraries, and so presumably would
be in the C++ libraries as well. It was also thought that implementations could provide
more efficient reference management then wrappers (such as boost::shared_ptr) around a noncopyable
thread concept. Analysis of whether or not these arguments would hold true don't appear to
bear them out. To illustrate the analysis we'll first provide pseudo-code illustrating the six
typical usage patterns of a thread object.</p>
<h3>1. Simple creation of a thread.</h3>
<pre>
void foo()
{
create_thread(&amp;bar);
}
</pre>
<h3>2. Creation of a thread that's later joined.</h3>
<pre>
void foo()
{
thread = create_thread(&amp;bar);
join(thread);
}
</pre>
<h3>3. Simple creation of several threads in a loop.</h3>
<pre>
void foo()
{
for (int i=0; i&lt;NUM_THREADS; ++i)
create_thread(&amp;bar);
}
</pre>
<h3>4. Creation of several threads in a loop which are later joined.</h3>
<pre>
void foo()
{
for (int i=0; i&lt;NUM_THREADS; ++i)
threads[i] = create_thread(&amp;bar);
for (int i=0; i&lt;NUM_THREADS; ++i)
threads[i].join();
}
</pre>
<h3>5. Creation of a thread whose ownership is passed to another object/method.</h3>
<pre>
void foo()
{
thread = create_thread(&amp;bar);
manager.owns(thread);
}
</pre>
<h3>6. Creation of a thread whose ownership is shared between multiple objects.</h3>
<pre>
void foo()
{
thread = create_thread(&amp;bar);
manager1.add(thread);
manager2.add(thread);
}
</pre>
<p>Of these usage patterns there's only one that requires reference management (number 6).
Hopefully it's fairly obvious that this usage pattern simply won't occur as often as the
other usage patterns. So there really isn't a "typical need" for a thread concept, though
there is some need.</p>
<p>Since the need isn't typical we must use different criteria for deciding on either a
thread_ref or thread design. Possible criteria include ease of use and performance. So let's
analyze both of these carefully.</p>
<p>With ease of use we can look at existing experience. The standard C++ objects that
represent a system resource, such as std::iostream, are noncopyable, so we know that C++
programmers must at least be experienced with this design. Most C++ developers are also
used to smart pointers such as boost::shared_ptr, so we know they can at least adapt to
a thread_ref concept with little effort. So existing experience isn't going to lead us
to a choice.</p>
<p>The other thing we can look at is how difficult it is to use both types for the six usage
patterns above. If we find it overly difficult to use a concept for any of the usage patterns
there would be a good argument for choosing the other design. So we'll code all six usage
patterns using both designs.</p>
<h3>1.</h3>
<pre>
void foo()
{
thread thrd(&amp;bar);
}
void foo()
{
thread_ref thrd = create_thread(&amp;bar);
}
</pre>
<h3>2.</h3>
<pre>
void foo()
{
thread thrd(&amp;bar);
thrd.join();
}
void foo()
{
thread_ref thrd =
create_thread(&amp;bar);thrd-&gt;join();
}
</pre>
<h3>3.</h3>
<pre>
void foo()
{
for (int i=0; i&lt;NUM_THREADS; ++i)
thread thrd(&amp;bar);
}
void foo()
{
for (int i=0; i&lt;NUM_THREADS; ++i)
thread_ref thrd = create_thread(&amp;bar);
}
</pre>
<h3>4.</h3>
<pre>
void foo()
{
std::auto_ptr&lt;thread&gt; threads[NUM_THREADS];
for (int i=0; i&lt;NUM_THREADS; ++i)
threads[i] = std::auto_ptr&lt;thread&gt;(new thread(&amp;bar));
for (int i= 0; i&lt;NUM_THREADS;
++i)threads[i]-&gt;join();
}
void foo()
{
thread_ref threads[NUM_THREADS];
for (int i=0; i&lt;NUM_THREADS; ++i)
threads[i] = create_thread(&amp;bar);
for (int i= 0; i&lt;NUM_THREADS;
++i)threads[i]-&gt;join();
}
</pre>
<h3>5.</h3>
<pre>
void foo()
{
thread thrd* = new thread(&amp;bar);
manager.owns(thread);
}
void foo()
{
thread_ref thrd = create_thread(&amp;bar);
manager.owns(thrd);
}
</pre>
<h3>6.</h3>
<pre>
void foo()
{
boost::shared_ptr&lt;thread&gt; thrd(new thread(&amp;bar));
manager1.add(thrd);
manager2.add(thrd);
}
void foo()
{
thread_ref thrd = create_thread(&amp;bar);
manager1.add(thrd);
manager2.add(thrd);
}
</pre>
<p>This shows the usage patterns being nearly identical in complexity for both designs.
The only actual added complexity occurs because of the use of operator new in (4), (5)
and (6) and the use of std::auto_ptr and boost::shared_ptr in (4) and (6) respectively.
However, that's not really much added complexity, and C++ programmers are used to using
these idioms any way. Some may dislike the presence of operator new in user code,
but this can be eliminated by proper design of higher level concepts, such as the
boost::thread_group class that simplifies example (4) down to:</p>
<pre>
void foo()
{
thread_group threads;
for (int i=0; i&lt;NUM_THREADS; ++i)
threads.create_thread(&amp;bar);
threads.join_all();
}
</pre>
<p>So ease of use is really a wash and not much help in picking a design.</p>
<p>So what about performance? If you look at the above code examples we can analyze
the theoretical impact to performance that both designs have. For (1) we can see that
platforms that don't have a ref-counted native thread type (POSIX, for instance) will
be impacted by a thread_ref design. Even if the native thread type is ref-counted there
may be an impact if more state information has to be maintained for concepts foreign
to the native API, such as clean up stacks for Win32 implementations. For (2) the
performance impact will be identical to (1). The same for (3). For (4) things get a
little more interesting and we find that theoretically at least the thread_ref may
perform faster since the thread design requires dynamic memory allocation/deallocation.
However, in practice there may be dynamic allocation for the thread_ref design as well,
it will just be hidden from the user. As long as the implementation has to do dynamic
allocations the thread_ref loses again because of the reference management. For (5)
we see the same impact as we do for (4). For (6) we still have a possible impact
to the thread design because of dynamic allocation but thread_ref no longer suffers
because of it's reference management, and in fact, theoretically at least, the thread_ref
may do a better job of managing the references. All of this indicates that thread wins
for (1), (2) and (3), with (4) and (5) the winner depends on the implementation and the platform
but the thread design probably has a better chance, and with (6) it will again
depend on the implementation and platform but this time we favor thread_ref slightly.
Given all of this it's a narrow margin, but the thread design prevails.</p>
<p>Given this analysis, and the fact that noncopyable objects for system resources are
the normal designs that C++ programmers are used to dealing with, the Boost.Threads
library has gone with a noncopyable design.</p>
<h2>Rationale for not providing <i><a name="Events">Event</a> Variables</i></h2>
<p><i>Event variables </i>are simply far too error-prone.&nbsp; <a href="condition.html">Condition
variables</a> are a much safer alternative.</p>
<p>[Note that Graphical User Interface <i>events</i> are a different concept,
and are not what is being discussed here.]</p>
<p>Event variables were one of the first synchronization primitives. They are
still used today, for example, in the native Windows multithreading API.</p>
<p>Yet both respected computer science researchers and experienced
multithreading practitioners believe event variables are so inherently
error-prone that they should never be used, and thus should not be part of a
multithreading library.</p>
<p>Per Brinch Hansen <a href="bibliography.html#Brinch-Hansen-73">[Brinch Hansen
73]</a> analyzed event variables in some detail, pointing out [emphasis his]
that &quot;<i>event operations force the programmer to be aware of the relative
speeds of the sending and receiving processes</i>&quot;. His summary:</p>
<blockquote>
<p>We must therefore conclude that event variables of the previous type are
impractical for system design. <i>The effect of an interaction between two
processes must be independent of the speed at which it is carried out.</i></p>
</blockquote>
<p>Experienced programmers using the Windows platform today
report that event variables are a continuing source of errors, even after previous
bad experiences caused them to be very careful in their use of event
variables.&nbsp; Overt problems can be avoided, for example, by teaming the
event variable with a mutex, but that may just convert a <a href="definitions.html#Race condition">race
condition</a> into another problem, such as excessive resource use. One of the most
distressing aspects of the experience reports is the claim that many defects are
latent.&nbsp; That is, the programs appear to work correctly, but contain
hidden timing dependencies which will cause them to fail when environmental
factors or usage patterns change, altering relative thread timings.</p>
<p>The decision to exclude event variables from Boost.Threads has been
surprising to some Windows programmers.&nbsp; They have written programs which
work using event variables, and wonder what the problem is.&nbsp; It seems
similar to the &quot;goto considered harmful&quot; controversy of 30 years ago.
It isn't that events, like gotos, can't be made to work, but rather that
virtually all programs using alternatives will be easier to write, debug,
read, maintain, and be less likely to contain latent defects.</p>
<p>[Rationale provided by Beman Dawes]</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <A href="mailto:williamkempf@hotmail.com">William E. Kempf</A>
2001 all rights reserved.</i></p>
</body>
</html>

327
doc/recursive_mutex.html Normal file
View File

@@ -0,0 +1,327 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, recursive_mutex</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">recursive_mutex<br>
recursive_try_mutex<br>
recursive_timed_mutex</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#recursive_mutex Synopsis">Class recursive_mutex Synopsis</a><br>
<a href="#recursive_mutex Members">Class recursive_mutex Members</a><br>
<a href="#recursive_try_mutex Synopsis">Class recursive_try_mutex Synopsis</a><br>
<a href="#recursive_try_mutex Members">Class recursive_try_mutex Members</a><br>
<a href="#recursive_timed_mutex Synopsis">Class recursive_timed_mutex Synopsis</a><br>
<a href="#recursive_timed_mutex Members">Class recursive_timed_mutex Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <code>recursive_mutex</code>, <code>recursive_try_mutex</code> and
<code>recursive_timed_mutex</code> classes define full featured models of the
<a href="mutex_concept.html#Mutex">Mutex</a>,
<a href="mutex_concept.html#TryMutex">TryMutex</a> and
<a href="mutex_concept.html#TimedMutex">TimedMutex</a> concepts with recursive locking
semantics. These types should be used to synchronize access to shared resources
when recursive locking by a single thread is likely to occur. A good example for this
is when a class supplies "internal synchronization" to ensure
<a href="definitions.html#Thread-safe">thread-safety</a> and a function of the class
may have to call other functions of the class which also attempt to lock the mutex.
For recursive locking mechanics, see <a href="mutex.html">mutexes</a>.
<p>Each class supplies one or more typedefs for lock types which model matching
lock concepts. For the best possible performance you should use the mutex class that
supports the minimum set of lock types that you need.</p>
<table border="1" cellpadding="5">
<tr>
<td><b>Mutex Class</b></td>
<td><b>Lock name</b></td>
<td><b>Implementation defined Lock Type</b></td>
<td><b>Lock Concept</b></td>
</tr>
<tr>
<td valign="top"><a href="#recursive_mutex Synopsis"><code>recursive_mutex</code></a></td>
<td valign="middle"><code>scoped_lock</code></td>
<td valign="middle"><a href="scoped_lock.html"><code>detail::thread::scoped_lock&lt;recursive_mutex&gt;</code></a></td>
<td valign="middle"><a href="lock_concept.html#ScopedLock">ScopedLock</a></td>
</tr>
<tr>
<td valign="top"><code><a href="#recursive_try_mutex Synopsis">recursive_try_mutex</a></code></td>
<td valign="middle"><code>scoped_lock<br>
scoped_try_lock</code></td>
<td valign="middle"><a href="scoped_lock.html"><code>detail::thread::scoped_lock&lt;recursive_try_mutex&gt;<br>
</code></a><code><a href="scoped_try_lock.html">detail::thread::scoped_try_lock&lt;recursive_try_mutex&gt;</a></code></td>
<td valign="middle"><a href="lock_concept.html#ScopedLock">ScopedLock</a><br>
<a href="lock_concept.html#ScopedTryLock">ScopedTryLock</a></td>
</tr>
<tr>
<td valign="top"><code><a href="#recursive_timed_mutex Synopsis">recursive_timed_mutex</a></code> </td>
<td valign="middle"><code>scoped_lock<br>
scoped_try_lock<br>
scoped_timed_lock</code></td>
<td valign="middle"><a href="scoped_lock.html"><code>detail::thread::scoped_lock&lt;recursive_timed_mutex&gt;</code></a><br>
<a href="scoped_try_lock.html"><code>detail::thread::scoped_try_lock&lt;recursive_timed_mutex&gt;</code></a><br>
<a href="scoped_timed_lock.html"><code>detail::thread::scoped_timed_lock&lt;recursive_timed_mutex&gt;</code></a></td>
<td valign="middle"><a href="lock_concept.html#ScopedLock">ScopedLock</a><br>
<a href="lock_concept.html#ScopedTryLock">ScopedTryLock</a><br>
<a href="lock_concept.html#ScopedTimedLock">ScopedTimedLock</a></td>
</tr>
</table>
<p>The <code>recursive_mutex</code>, <code>recursive_try_mutex</code> and
<code>recursive_timed_mutex</code> employ a <code>Recursive</code>
<a href="mutex_concept.html#LockingStrategies">locking strategy</a>, so attempts to
recursively lock them succeed and an internal "lock count" is maintained. Attempts
to unlock them by a thread that does not own a lock on them will result in a
<a href="lock_error.html">lock_error</a> exception being thrown.</p>
<p>The <code>recursive_mutex</code>, <code>recursive_try_mutex</code> and
<code>recursive_timed_mutex</code> leave the
<a href="mutex_concept.html#SchedulingPolicies">scheduling policy</a> as
<code>Unspecified</code>. Programmers should assume that threads waiting for a lock on
objects of these types acquire the lock in a random order, even though the specific
behavior for a given platform may be different.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/recursive_mutex.hpp">&lt;boost/thread/recursive_mutex.hpp&gt;</a>
</pre>
<h2>Class <a name="recursive_mutex Synopsis"> recursive_mutex Synopsis</a></h2>
<hr>
<pre>
namespace boost
{
class recursive_mutex : private <a href="../../utility/utility.htm">boost::noncopyable</a> // Exposition only.
// Class recursive_mutex meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_lock;
recursive_mutex();
~recursive_mutex();
};
}
</pre>
<h2>Class <a name="recursive_mutex Members">recursive_mutex Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
recursive_mutex();
</pre>
<p><b>Postconditions: </b><code>*this</code> is in the unlocked state.</p>
<hr>
<h3>Destructor</h3>
<pre>
~recursive_mutex();
</pre>
<p><b>Requires:</b> <code>*this</code> is in the unlocked state.</p>
<p><b>Effects:</b> Destroys <code>*this</code>.</p>
<p><b>Dangers:</b> Destruction of a locked mutex is a serious programming error
resulting in undefined behavior such as a program crash..</p>
<hr>
<h2>
Class <a name="recursive_try_mutex Synopsis">recursive_try_mutex Synopsis</a>
</h2>
<pre>
namespace boost
{
class recursive_try_mutex : private boost::noncopyable // Exposition only.
// Class recursive_try_mutex meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_lock;
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_try_lock;
recursive_try_mutex();
~recursive_try_mutex();
};
}
</pre>
<h2>Class <a name="recursive_try_mutex Members">recursive_try_mutex Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
recursive_try_mutex();
</pre>
<p><b>Postconditions: </b><code>*this</code> is in the unlocked state.</p>
<hr>
<h3>Destructor</h3>
<pre>
~recursive_try_mutex();
</pre>
<p><b>Requires:</b> <code>*this</code> is in the unlocked state.</p>
<p><b>Effects:</b> Destroys <code>*this</code>.</p>
<p><b>Dangers:</b> Destruction of a locked mutex is a serious programming error
resulting in undefined behavior such as a program crash..</p>
<hr>
<h2>
Class <a name="recursive_timed_mutex Synopsis">recursive_timed_mutex Synopsis</a>
</h2>
<pre>
namespace boost
{
class recursive_timed_mutex : private boost::noncopyable // Exposition only.
// Class recursive_timed_mutex meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_lock;
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_try_lock;
typedef <i>[implementation defined; see <a href="#Introduction">Introduction</a>]</i> scoped_timed_lock;
recursive_timed_mutex();
~recursive_timed_mutex();
};
}
</pre>
<h2>
Class <a name="recursive_timed_mutex Members">recursive_timed_mutex Members</a>
</h2>
<hr>
<h3>Constructor</h3>
<pre>
recursive_timed_mutex();
</pre>
<p><b>Postconditions: </b><code>*this</code> is in the unlocked state.</p>
<hr>
<h3>Destructor</h3>
<pre>
~recursive_timed_mutex();
</pre>
<p><b>Requires:</b> <code>*this</code> is in the unlocked state.</p>
<p><b>Effects:</b> Destroys <code>*this</code>.</p>
<p><b>Dangers:</b> Destruction of a locked mutex is a serious programming error
resulting in undefined behavior such as a program crash..</p>
<hr>
<h2><a name="Example">Example</a> Usage</h2>
<pre>
#include <a href="../../../boost/thread/recursive_mutex.hpp">&lt;boost/thread/recursive_mutex.hpp&gt;</a>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
#include &lt;iostream&gt;
class counter
{
public:
counter() : count(0) { }
int add(int val) {
boost::recursive_mutex::scoped_lock scoped_lock(mutex);
count += val;
return count;
}
int increment() {
boost::recursive_mutex::scoped_lock scoped_lock(mutex);
return add(1);
}
private:
boost::recursive_mutex mutex;
int count;
};
counter c;
void change_count(void*)
{
std::cout &lt;&lt; &quot;count == &quot; &lt;&lt; c.increment() &lt;&lt; std::endl;
}
int main(int, char*[])
{
const int num_threads=4;
boost::thread_group threads;
for (int i=0; i &lt; num_threads; ++i)
threads.create_thread(&amp;change_count, 0);
threads.join_all();
return 0;
}
</pre>
<p>The output is:</p>
<pre>
count == 1
count == 2
count == 3
count == 4
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

134
doc/removed_atomic_t.html Normal file
View File

@@ -0,0 +1,134 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<title>Boost.Threads, atomic_t</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">atomic_t</h2>
</td>
</tr>
</table>
<hr>
<h2>Header</h2>
<p>The <tt>atomic_t</tt> class defines an "atomic integer" type. This class should be used
to perform thread safe operations on an integral type with out the overhead of locks. Only
a limited set of integer operations are available with an <tt>atomic_t</tt> instance.</p>
<pre>
#include &lt;boost/thread/atomic.hpp&gt;
</pre>
<h2>Public Interface</h2>
<pre>
class atomic_t
{
public:
typedef <b>implementation defined</b> value_type;
explicit atomic_t(value_type val=0);
};
atomic_t::value_type read(const atomic_t&amp; x);
atomic_t::value_type increment(atomic_t&amp; x);
atomic_t::value_type decrement(atomic_t&amp; x);
atomic_t::value_type swap(atomic_t&amp; x, atomic_t::value_type y);
atomic_t::value_type compare_swap(atomic_t&amp; x, atomic_t::value_type y, atomic_t::value_type z);
</pre>
<h3>Constructor</h3>
<pre>
atomic_t(atomic_t::value_type val=0);
</pre>
<p>Constructs an <tt>atomic_t</tt> and sets its value to <tt>val</tt>.</p>
<h3>read</h3>
<pre>
atomic_t::value_type read(const atomic_t&amp; x);
</pre>
<p>Gets the current value of <tt>x</tt>.</p>
<h3>increment</h3>
<pre>
atomic_t::value_type increment(atomic_t&amp; x);
</pre>
<p>Increments <tt>x</tt> and returns a value <tt>&lt; 0</tt> if the result is less than 0,
<tt>&gt; 0</tt> if the result is greater than 0 and <tt>== 0</tt> if the result is equal to
0.</p>
<h3>decrement</h3>
<pre>
atomic_t::value_type decrement(atomic_t&amp; x);
</pre>
<p>Decrements <tt>x</tt> and returns a value <tt>&lt; 0</tt> if the result is less than 0,
<tt>&gt; 0</tt> if the result is greater than 0 and <tt>== 0</tt> if the result is equal to
0.</p>
<h3>swap</h3>
<pre>
atomic_t::value_type swap(atomic_t&amp; x, atomic_t::value_type y);
</pre>
<p>Assigns the value of <tt>y</tt> to <tt>x</tt> and returns the value of <tt>x</tt> prior
to the swap.</p>
<h3>compare_swap</h3>
<pre>
atomic_t::value_type compare_swap(atomic_t&amp; x, atomic_t::value_type y, atomic_t::value_type z);
</pre>
<p>Compares the value of <tt>z</tt> to the value of <tt>x</tt> and if equal sets the value of
<tt>x</tt> to the value of <tt>y</tt> and returns the value of <tt>x</tt> prior to the swap.</p>
<h2>Example Usage</h2>
<pre>
#include &lt;boost/thread/atomic.hpp&gt;
#include &lt;boost/test/test_tools.hpp&gt;
int test_main(int, char*[])
{
boost::atomic_t a;
BOOST_TEST_VERIFY(boost::read(a) == 0);
BOOST_TEST_VERIFY(boost::increment(a) &gt; 0);
BOOST_TEST_VERIFY(boost::decrement(a) == 0);
BOOST_TEST_VERIFY(boost::swap(a, 1) == 0);
BOOST_TEST_VERIFY(boost::swap(a, 2, 0) == 1);
BOOST_TEST_VERIFY(boost::read(a) == 1);
}
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

213
doc/scoped_lock.html Normal file
View File

@@ -0,0 +1,213 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, scoped_lock</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">scoped_lock</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#Synopsis">Synopsis</a><br>
<a href="#Members">Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>This class template defines a generic lock type which meets the
<a href="lock_concept.html#ScopedLock">ScopedLock</a> requirements. The
<a href="mutex.html">mutex</a>, <a href="mutex.html">try_mutex</a>,
<a href="mutex.html">timed_mutex</a>, <a href="recursive_mutex.html">recursive_mutex</a>,
<a href="recursive_mutex.html">recursive_try_mutex</a> and
<a href="recursive_mutex.html">recursive_timed_mutex</a> classes all use this template
to define their <code>scoped_lock</code> types.</p>
<p>Like all the <b>Boost.Threads</b> <a href="lock_concept.html">lock models</a>,
<code>scoped_lock</code> objects are meant to be short-lived. Objects of the class
are not <a href="definitions.html#thread-safe">thread-safe</a>, and so should not be
shared between threads.</p>
<p>Class <code> scoped_lock</code> follows the &quot;resource acquisition is
initialization&quot; idiom <a href="bibliography.html#Stroustrup-00">[Stroustrup
00 14.4.1]</a> and is a realization of the &quot;Scoped Locking Pattern&quot;
<a href="bibliography.html#Schmidt-00">[Schmidt-00]</a>. Thus the usage is to let the
constructor do the locking, and then let the destructor do the unlocking automatically at
the end of the enclosing scope. The lock() and unlock() members are usually not
explicitly called, but are provided to allow for complex overlapping locks of multiple
mutexes.</p>
<p>The type used to instantiate the class must meet the
<a href="mutex_concept.html#Mutex">Mutex</a> requirements.</p>
<p>Although this class is an implementation detail, it is publicly documented here because
of its importance.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/detail/lock.hpp">&lt;boost/thread/detail/lock.hpp&gt;</a>
<i>This header is usually not included directly by programmers
because it is supplied by <a href="../../../boost/thread/mutex.hpp">&lt;boost/thread/mutex.hpp&gt;</a> or
<a href="../../../boost/thread/recursive_mutex.hpp">&lt;boost/thread/recursive_mutex.hpp&gt;</a></i>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost { namespace detail { namespace thread {
template &lt;typename Mutex&gt;
class scoped_lock : private <a href="../../utility/utility.htm#Class noncopyable">boost::noncopyable</a> // Exposition only.
// Class scoped_lock meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef Mutex mutex_type;
explicit scoped_lock(Mutex&amp; mx, bool initially_locked=true);
~scoped_lock();
void lock();
void unlock();
operator const void*() const;
bool locked() const;
};
} // namespace thread
} // namespace detail
} // namespace boost
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
explicit scoped_lock(Mutex&amp; mx, bool initially_locked=true);
</pre>
<p><b>Effects:</b> Associates mutex <code>mx</code> with <code>*this</code>.
If <code>initially_locked</code> is <code>true,</code> calls <code>lock()</code>.</p>
<hr>
<h3>Destructor</h3>
<pre>
~scoped_lock();
</pre>
<p><b>Effects:</b> If <code> locked()</code>, calls <code>unlock()</code>. Destroys
<code>*this</code>.</p>
<hr>
<h3>lock</h3>
<pre>
void lock();
</pre>
<p><b>Effects:</b> If the associated mutex is already locked by another lock in the
current thread, the effects depend on the locking strategy of the associated mutex, as
shown in the following table:</p>
<table border="1" cellpadding="5">
<tr>
<td><i><a href="mutex_concept.html#LockingStrategies">Locking Strategy</a><br>
of associated mutex</i></td>
<td><i>Effect if associated mutex is already locked by the current thread</i></td>
</tr>
<tr>
<td>Recursive</td>
<td>As if an additional lock were added to the mutex.</td>
</tr>
<tr>
<td>Checked</td>
<td>Throws <a href="lock_error.html">lock_error</a>.</td>
</tr>
<tr>
<td>Unchecked</td>
<td>Undefined behavior [<a href="bibliography.html#ISO">ISO</a> 1.3.12] (but
typically, <a href="definitions.html#Deadlock">deadlock</a>.)</td>
</tr>
</table>
<p>If the associated mutex is already locked by some other thread, places the
current thread in the <a href="definitions.html#State">Blocked</a> state until
the associated mutex is unlocked, after which the current thread is placed in
the <a href="definitions.html#State">Ready</a> state, eventually to be returned
to the <a href="definitions.html#State">Running</a> state.
<p><b>Postcondition:</b> locked()
<p><b>Throws:</b> <a href="lock_error.html">lock_error</a> if <code>locked()</code> or
as indicated in <b>Effects</b>.</p>
<hr>
<h3>unlock</h3>
<pre>
void unlock();
</pre>
<p><b>Effects: </b>Unlocks the associated mutex.</p>
<p><b>Throws:</b> <a href="lock_error.html">lock_error</a> if <code>!locked()</code>.</p>
<hr>
<h3>const void* Conversion</h3>
<pre>
operator const void*() const;
</pre>
<p><b>Returns:</b> If the associated mutex is currently locked, a value convertible to
<code>true</code>, else a value convertible to <code>false</code>.</p>
<p><b>Rationale:</b> A <code>const void*</code> conversion is considered safer than a
conversion to <code>bool</code>.</p>
<hr>
<h3>locked</h3>
<pre>
bool locked() const;
</pre>
<p><b>Returns:</b> <code>this-&gt;operator const void*() != 0</code>.</p>
<hr>
<h2><a name="Example">Example</a> Usage</h2>
<p>See the example given in the documentation for the <a href="mutex.html">mutex</a>
class.</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

261
doc/scoped_timed_lock.html Normal file
View File

@@ -0,0 +1,261 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, scoped_timed_lock</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">scoped_timed_lock</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#Synopsis">Synopsis</a><br>
<a href="#Members">Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>This class template defines a generic lock type which meets the
<a href="lock_concept.html#ScopedTimedLock">ScopedTimedLock</a> requirements. The
<a href="mutex.html">timed_mutex</a> and
<a href="recursive_mutex.html">recursive_timed_mutex</a> classes use this template to
define their <code>scoped_timed_lock</code> types.</p>
<p>Like all the <b>Boost.Threads</b> <a href="lock_concept.html">lock models</a>,
<code>scoped_timed_lock</code> objects are meant to be short-lived. Objects of the
class are not <a href="definitions.html#thread-safe">thread-safe</a>, and so should
not be shared between threads.</p>
<p>Class <code>scoped_timed_lock</code> follows the &quot;resource acquisition is
initialization&quot; idiom <a href="bibliography.html#Stroustrup-00">[Stroustrup
00 14.4.1]</a> and is a realization of the &quot;Scoped Locking Pattern&quot;
<a href="bibliography.html#Schmidt-00">[Schmidt-00]</a>. Thus the usage is to let the
constructor do the locking, and then let the destructor do the unlocking automatically
at the end of the enclosing scope. The lock() and unlock() members are usually not
explicitly called, but are provided to allow for complex overlapping locks of multiple
mutexes.</p>
<p>The type used to instantiate the class must meet the
<a href="mutex_concept.html#TimedMutex">TimedMutex</a> requirements.</p>
<p>Although this class is an implementation detail, it is publicly documented here because
of its importance.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/detail/lock.hpp">&lt;boost/thread/detail/lock.hpp&gt;</a>
<i>This header is usually not included directly by programmers
because it is supplied by <a href="../../../boost/thread/mutex.hpp">&lt;boost/thread/mutex.hpp&gt;</a> or
<a href="../../../boost/thread/recursive_mutex.hpp">&lt;boost/thread/recursive_mutex.hpp&gt;</a></i>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost { namespace detail { namespace thread {
template &lt;typename TimedMutex&gt;
class scoped_timed_lock : private <a href="../../utility/utility.htm#Class noncopyable">boost::noncopyable</a> // Exposition only.
// Class scoped_timed_lock meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef TimedMutex mutex_type;
scoped_timed_lock(TimedMutex&amp; mx, const boost::xtime&amp; xt);
scoped_timed_lock(TimedMutex&amp; mx, bool initially_locked);
~scoped_timed_lock();
void lock();
bool timed_lock(const xtime&amp; xt);
void unlock();
operator const void*() const;
};
} // namespace thread
} // namesapce detail
} // namespace boost
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
scoped_timed_lock(TimedMutex&amp; mx, const <a href="xtime.html">xtime</a>&amp; xt);
</pre>
<p><b>Effects:</b> Associates mutex <code>mx</code> with <code>*this</code>.
Calls <code>timed_lock</code>( <code>xt</code>)</p>
<hr>
<pre>
scoped_timed_lock(TimedMutex&amp; mx, bool initially_locked);
</pre>
<p><b>Effects:</b> Associates mutex <code>mx</code> with <code>*this</code>.
If <code>initially_locked</code> is <code>true</code>, calls <code>lock()</code>.</p>
<hr>
<h3>Destructor</h3>
<pre>
~scoped_timed_lock();
</pre>
<p><b>Effects:</b> If <code> locked()</code>, calls <code>unlock()</code>. Destroys
<code>*this</code>.</p>
<hr>
<h3>lock</h3>
<pre>
void lock();
</pre>
<p><b>Effects:</b> If the associated mutex is already locked by another lock in the
current thread, the effects depend on the locking strategy of the associated mutex, as
shown in the following table:</p>
<table border="1" cellpadding="5">
<tr>
<td><i><a href="mutex_concept.html#LockingStrategies">Locking Strategy</a><br>
of associated mutex</i></td>
<td><i>Effect if associated mutex is already locked by the current thread</i></td>
</tr>
<tr>
<td>Recursive</td>
<td>As if an additional lock were added to the mutex.</td>
</tr>
<tr>
<td>Checked</td>
<td>Throws <a href="lock_error.html">lock_error</a>.</td>
</tr>
<tr>
<td>Unchecked</td>
<td>Undefined behavior [<a href="bibliography.html#ISO">ISO</a> 1.3.12] (but
typically, <a href="definitions.html#Deadlock">deadlock</a>.)</td>
</tr>
</table>
<p>If the associated mutex is already locked by some other thread, places the
current thread in the <a href="definitions.html#State">Blocked</a> state until
the associated mutex is unlocked, after which the current thread is placed in
the <a href="definitions.html#State">Ready</a> state, eventually to be returned
to the <a href="definitions.html#State">Running</a> state. Places the associated
mutex in the locked state.
<p><b>Throws:</b> <a href="lock_error.html">lock_error</a> if <code>locked()</code> or
as indicated in <b>Effects</b>.</p>
<hr>
<h3>timed_lock</h3>
<pre>
bool timed_lock(const <a href="xtime.html">xtime</a>&amp; xt);
</pre>
<p><b>Effects:</b> Same as <code>lock()</code>, except that if xt is reached,
places the current thread in the <a href="definitions.html#State">Ready</a>
state without further ado.</p>
<p><b>Returns:</b> <code>locked()</code>.</p>
<p><b>Throws:</b> <a href="lock_error.html">lock_error</a> if <code>locked()</code> or
as indicated in <b>Effects</b>.</p>
<hr>
<h3>unlock</h3>
<pre>
void unlock();
</pre>
<p><b>Effects: </b>Unlocks the associated mutex.</p>
<p><b>Throws:</b> <a href="lock_error.html">lock_error</a> if <code>!locked()</code>.</p>
<hr>
<h3>const void* Conversion</h3>
<pre>
operator const void*() const;
</pre>
<p><b>Returns:</b> If the associated mutex is currently locked, a value convertible to
<code>true</code>, else a value convertible to <code>false</code>.</p>
<p><b>Rationale:</b> A <code>const void*</code> conversion is considered safer than a
conversion to <code>bool</code>.</p>
<hr>
<h3>locked</h3>
<pre>
bool locked() const;
</pre>
<p><b>Returns:</b> <code>this-&gt;operator const void*() != 0</code>.</p>
<hr>
<h2><a name="Example">Example</a> Usage</h2>
<pre>
#include &lt;boost/thread/mutex.hpp&gt;
#include &lt;iostream&gt;
int main(int, char*[])
{
boost::timed_mutex mutex;
boost::xtime xt;
boost::get_xtime(&amp;xt, boost::TIME_UTC);
xt.sec += 1;
boost::mutex::scoped_timed_lock scope_timed_lock(mutex, xt);
if (scope_timed_lock.locked())
std::cout &lt;&lt; &quot;locked&quot; &lt;&lt; std::endl;
else
std::cout &lt;&lt; &quot;unlocked&quot; &lt;&lt; std::endl;
return 0;
}
</pre>
<p>The output is:</p>
<pre>
locked
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%B %d, %Y" startspan -->October 01, 2001<!--webbot bot="Timestamp" endspan i-checksum="30079" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

282
doc/scoped_try_lock.html Normal file
View File

@@ -0,0 +1,282 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, scoped_try_lock</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">scoped_try_lock</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#Synopsis">Synopsis</a><br>
<a href="#Members">Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>This class template defines a generic lock type which meets the
<a href="lock_concept.html#ScopedTryLock">ScopedTryLock</a> requirements. The
<a href="mutex.html">try_mutex</a>, <a href="mutex.html">timed_mutex</a>,
<a href="recursive_mutex.html">recursive_try_mutex</a> and
<a href="recursive_mutex.html">recursive_timed_mutex</a> classes use this template
to define their <code>scoped_try_lock</code> types.</p>
<p>Like all the <b>Boost.Threads</b> <a href="lock_concept.html">lock models</a>,
<code>scoped_try_lock</code> objects are meant to be short-lived. Objects of the
class are not <a href="definitions.html#thread-safe">thread-safe</a>, and
so should not be shared between threads.</p>
<p>Class <code> scoped_try_lock</code> follows the &quot;resource acquisition is
initialization&quot; idiom <a href="bibliography.html#Stroustrup-00">[Stroustrup
00 14.4.1]</a> and is a realization of the &quot;Scoped Locking Pattern&quot;
<a href="bibliography.html#Schmidt-00">[Schmidt-00]</a>. Thus the usage is to let the
constructor do the locking, and then let the destructor do the unlocking automatically at
the end of the enclosing scope. The lock() and unlock() members are usually not
explicitly called, but are provided to allow for complex overlapping locks of multiple
mutexes.</p>
<p>Although this class is an implementation detail, it is publicly documented here because
of its importance.</p>
<p>The type used to instantiate the class must meet the <a href="mutex_concept.html#TryMutex">TryMutex</a> requirements.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/detail/lock.hpp">&lt;boost/thread/detail/lock.hpp&gt;</a>
<i>This header is usually not included directly by programmers
because it is supplied by <a href="../../../boost/thread/mutex.hpp">&lt;boost/thread/mutex.hpp&gt;</a> or
<a href="../../../boost/thread/recursive_mutex.hpp">&lt;boost/thread/recursive_mutex.hpp&gt;</a></i>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost { namespace detail { namespace thread {
template &lt;typename TryMutex&gt;
class scoped_try_lock : private <a href="../../utility/utility.htm#Class noncopyable">boost::noncopyable</a> // Exposition only.
// Class scoped_try_lock meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
typedef TryMutex mutex_type;
explicit scoped_try_lock(TryMutex&amp; mx);
scoped_try_lock(TryMutex&amp; mx, bool initially_locked);
~scoped_try_lock();
void lock();
bool try_lock();
void unlock();
operator const void*() const;
};
} // namespace thread
} // namespace detail
} // namespace boost
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructors</h3>
<pre>
explicit scoped_try_lock(TryMutex&amp; mx);
</pre>
<p><b>Effects:</b> Associates mutex <code>mx</code> with <code>*this</code>.
Calls <code>try_lock()</code>.</p>
<hr>
<pre>
scoped_try_lock(TryMutex&amp; mx, bool initially_locked);
</pre>
<p><b>Effects:</b> Associates mutex <code>mx</code> with <code>*this</code>.
If <code>initially_locked</code> is <code>true,</code> calls <code>lock()</code>.</p>
<hr>
<h3>Destructor</h3>
<pre>
~scoped_try_lock();
</pre>
<p><b>Effects:</b> If <code>locked()</code>, calls <code>unlock()</code>. Destroys
<code>*this</code>.</p>
<hr>
<h3>lock</h3>
<pre>
void lock();
</pre>
<p><b>Effects:</b> If the associated mutex is already locked by another lock in the
current thread, the effects depend on the locking strategy of the associated mutex, as
shown in the following table:</p>
<table border="1" cellpadding="5" height="147">
<tr>
<td height="34"><i><a href="mutex_concept.html#LockingStrategies">Locking Strategy</a><br>
of associated mutex</i></td>
<td height="34"><i>Effect if associated mutex is already locked by the
current thread</i></td>
</tr>
<tr>
<td height="19">Recursive</td>
<td height="19">As if an additional lock were added to the mutex.</td>
</tr>
<tr>
<td height="19">Checked</td>
<td height="19">Throws <a href="lock_error.html">lock_error</a>.</td>
</tr>
<tr>
<td height="19">Unchecked</td>
<td height="19">Undefined behavior [<a href="bibliography.html#ISO">ISO</a> 1.3.12] (but
typically, <a href="definitions.html#Deadlock">deadlock</a>.)</td>
</tr>
</table>
<p>If the associated mutex is already locked by some other thread, places the
current thread in the <a href="definitions.html#State">Blocked</a> state until
the associated mutex is unlocked, after which the current thread is placed in
the <a href="definitions.html#State">Ready</a> state, eventually to be returned
to the <a href="definitions.html#State">Running</a> state. Places the associated
mutex in the locked state.</p>
<p><b>Throws:</b> <a href="lock_error.html">lock_error</a> if <code>locked()</code> or
as indicated in <b>Effects</b>.</p>
<hr>
<h3>try_lock</h3>
<pre>
bool try_lock();
</pre>
<p><b>Effects:</b> If the associated mutex is already locked by another lock in the
current thread, the effects depend on the locking strategy of the associated mutex, as
shown in the following table:</p>
<table border="1" cellpadding="5" height="147">
<tr>
<td height="34"><i><a href="mutex_concept.html#LockingStrategies">Locking Strategy</a><br>
of associated mutex</i></td>
<td height="34"><i>Effect if associated mutex is already locked by the
current thread</i></td>
</tr>
<tr>
<td height="19">Recursive</td>
<td height="19">As if an additional lock were added to the mutex.</td>
</tr>
<tr>
<td height="19">Checked</td>
<td height="19">Throws <a href="lock_error.html">lock_error</a>.</td>
</tr>
<tr>
<td height="19">Unspecified</td>
<td height="19">Undefined behavior [<a href="bibliography.html#ISO">ISO</a> 1.3.12] (but
typically, <a href="definitions.html#Deadlock">deadlock</a>.)</td>
</tr>
</table>
<p>If the associated mutex is not already locked by some other thread, locks the
associated mutex and returns true, else returns false.</p>
<p><b>Returns:</b> See effects.</p>
<p><b>Throws:</b> <a href="lock_error.html">lock_error</a> if <code>locked()</code> or
as indicated in <b>Effects</b>.</p>
<hr>
<h3>unlock</h3>
<pre>
void unlock();
</pre>
<p><b>Effects: </b>Unlocks the associated mutex.</p>
<p><b>Throws:</b> <a href="lock_error.html">lock_error</a> if <code>!locked()</code>.</p>
<hr>
<h3>const void* Conversion</h3>
<pre>
operator const void*() const;
</pre>
<p><b>Returns:</b> If the associated mutex is currently locked, a value convertible to
<code>true</code>, else a value convertible to <code>false</code>.</p>
<p><b>Rationale:</b> A <code>const void*</code> conversion is considered safer
than a conversion to <code>bool</code>.</p>
<hr>
<h3>locked</h3>
<pre>
bool locked() const;
</pre>
<p><b>Returns:</b> <code>this-&gt;operator const void*() != 0</code>.</p>
<hr>
<h2><a name="Example">Example</a> Usage</h2>
<pre>
#include <a href="../../../boost/thread/mutex.hpp">&lt;boost/thread/mutex.hpp&gt;</a>
#include &lt;iostream&gt;
int main(int, char*[])
{
boost::mutex mutex;
boost::mutex::try_lock lock(mutex);
if (lock)
std::cout &lt;&lt; &quot;locked&quot; &lt;&lt; std::endl;
else
std::cout &lt;&lt; &quot;unlocked&quot; &lt;&lt; std::endl;
return 0;
}
</pre>
<p>The output is:</p>
<pre>
locked
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

226
doc/semaphore.html Normal file
View File

@@ -0,0 +1,226 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, semaphore</title>
</head>
<body bgcolor="#ffffff" link="#0000ff" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><IMG height=86 alt="C++ Boost" src="../../../c++boost.gif" width=277></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">semaphore</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#Synopsis">Synopsis</a><br>
<a href="#Members">Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <tt>semaphore</tt> class defines a classic synchronization primitive invented by the
Dutch computer scientist Edsger W. Dijkstra. A semaphore manages an internal counter. This
counter never goes below zero, or above a specified maximum value. When calling
<tt>semaphore::down</tt> the calling thread will block until the value is non-zero and then
decrement the value in a single atomic operation. When calling <tt>semaphore::up</tt> the
calling thread will increment the value in a single atomic operation, failing if the value has
already reached the specified maximum.</p>
<p><b>Rationale:</b> The semaphore is the simplest synchronization primitive available and is generally the
primitive used to build other synchronization concepts at some level of implementation. For this
reason <b>Boost.Threads</b> defines the <tt>semaphore</tt> type in the classic form. This simplifies
usage and implementation, but it means that the interface is not as safe as other <b>Boost.Threads</b>
interfaces.</p>
<p><b><a name="Danger">Danger</a>:</b> Unlike the <A href="mutex_concept.html">mutex models</a> supplied by <b>Boost.Threads,</b>
there is no <A href="lock_concept.html">lock_concept</a> for the semaphore to help ensure proper
usage. Great care must be taken when using a <tt>semaphore</tt> object to ensure
<a href="definitions.html#Deadlock"> deadlock</a> or <a href="definitions.html#Race condition">race conditions</a> do not occur.&nbsp;</p>
<p>The dangers are spelled out by <a href="bibliography.html#Andrews-83">[Andrews-83]</a>
(function names updated, see historical note below):&nbsp;</p>
<blockquote>
<p>Although semaphores can be used to program almost any kind of synchronization,
<b>down()</b> and <b>up()</b> are rather unstructured primitives, and so it is easy to err when using them. Execution of each critical section must begin with a
<b>down()</b> and end with a <b>up()</b> (on the same semaphore). Omitting a <b>down()</b>
or <b>up()</b>, or accidentally coding a <b>down()</b> on one semaphore and a <b>up()</b>
on another can have disastrous effects, since mutually exclusive execution would no longer be ensured. Also, when using semaphores, a programmer can forget to include in critical sections all statements that reference shared objects. This, too, could destroy the mutual exclusion required within critical sections. A second difficulty with using semaphores is that both condition synchronization and mutual exclusion are programmed using the same pair of primitives. This makes it difficult to identify the purpose of a given
<b>down()</b> or <b>up()</b> operation without looking at the other operations on the corresponding semaphore. Since mutual exclusion and condition synchronization are distinct concepts, they should have distinct notations.</p>
</blockquote>
<p><b>Historical note: </b>Dijkstra's original name for <b>down()</b> was <b>P</b>
(short for the Dutch &quot;passeren&quot;, &quot;to pass&quot;), and for <b>up()</b>
was <b>V</b> (short for the Dutch &quot;vrygeven&quot;, &quot;to release&quot;).</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/semaphore.hpp">&lt;boost/thread/semaphore.hpp&gt;</a>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost
{
class semaphore : private <a href="../../utility/utility.htm#Class noncopyable">boost::noncopyable</a> // Exposition only.
// Class semaphore meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
explicit semaphore(unsigned count=0, unsigned max=0);
~semaphore();
bool up(unsigned count=1, unsigned* prev=0);
void down();
bool down(const xtime&amp; xt);
private:
unsigned m_count; <i>exposition only [ISO 17.3.2.3/2]
</i> unsigned m_max; <i>exposition only [ISO 17.3.2.3/2]</i>
};
}
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
explicit semaphore(unsigned count=0, unsigned max=0);
</pre>
<p><b>Effects:</b> As if:</p>
<p><code>&nbsp;&nbsp;&nbsp;&nbsp; m_count = count;<br>
&nbsp;&nbsp;&nbsp;&nbsp; m_max = (max == 0 ? std::numeric_limits&lt;unsigned&gt;::max()
? max );</code></p>
<hr>
<h3>Destructor</h3>
<pre>
~semaphore();
</pre>
<p><b>Effects:</b> Destroys <code>*this</code>.</p>
<hr>
<h3>up</h3>
<pre>
bool up(unsigned count=1, unsigned* prev=0);
</pre>
<p><b>Effects:</b> As if:</p>
<p><code>&nbsp;&nbsp;&nbsp; unsigned ct;<br>
&nbsp;&nbsp;&nbsp; bool ret;<br>
&nbsp;&nbsp;&nbsp; { // as a single atomic operation:<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ct = m_count;<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; if (m_count == m_max) ret =
false;<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; else<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; {<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ret =
true;<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ++m_count;<br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; }<br>
&nbsp;&nbsp;&nbsp; }<br>
&nbsp;&nbsp;&nbsp; if (prev) *prev = m_count;<br>
&nbsp;&nbsp;&nbsp; return ret;</code></p>
<hr>
<h3>down</h3>
<pre>
void down();
</pre>
<p><b>Effects:</b> If <code>m_count == 0</code>, places the current thread in
the <a href="definitions.html#State">blocked</a> state until <code>m_count != 0</code>.
Finally, <code>--m_count</code>.<code>&nbsp;</code></p>
<hr>
<pre>
bool down(const <a href="xtime.html">xtime</a>&amp; xt);
</pre>
<p><b>Effects:</b> If <code>m_count == 0</code>, places the current thread in
the <a href="definitions.html#State">blocked</a> state until <code>m_count != 0</code>
or <code>xt</code> is reached. Finally, <code>--m_count</code>.<code>&nbsp;</code></p>
<p><b>Returns:</b> If xt was reached, true, else false.</p>
<hr>
<h2><a name="Example">Example</a> Usage</h2>
<pre>
#include <a href="../../../boost/thread/semaphore.hpp">&lt;boost/thread/semaphore.hpp&gt;</a>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
#include &lt;iostream&gt;
int global_data = 0;
boost::semaphore global_semaphore(1);
void change_global_data(void*)
{
global_semaphore.down();
++global_data;
std::cout &lt;&lt; &quot;global_data == &quot; &lt;&lt; global_data &lt;&lt; std::endl;
global_semaphore.up();
}
int main(int, char*[])
{
const int num_threads = 4;
boost::thread_group thrds;
for (int i=0; i &lt; num_threads; ++i)
thrds.create_thread(&amp;change_global_data, 0);
thrds.join_all();
return 0;
}
</pre>
<p>The output is:</p>
<pre>
global_data == 1
global_data == 2
global_data == 3
global_data == 4
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <A href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

4
doc/styles.css Normal file
View File

@@ -0,0 +1,4 @@
PRE
{
BACKGROUND-COLOR: lightcyan
}

254
doc/thread.html Normal file
View File

@@ -0,0 +1,254 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, Boost.Threads, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, thread</title>
</head>
<body bgcolor="#ffffff" link="#0000ff" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><IMG height=86 alt="C++ Boost" src="../../../c++boost.gif" width=277></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">Class thread</h2>
</td>
</tr>
</table>
<hr>
<p><A href="#Introduction">Introduction</A><br>
<A href="#Header">Header</A><br>
<A href="#Synopsis">Synopsis</A><br>
<A href="#Members">Members</A><br>
<A href="#Example">Example</A></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <code>thread</code> class represents threads of execution, and provides
the functionality to create and manage threads within the <b>Boost.Threads</b>
library. See <A href="definitions.html">Definitions</A> for a precise description of
"thread of execution", and for definitions of threading related terms and of thread
states such as "blocked".</p>
<p>A thread of execution has an initial function. For the program's
initial thread, the initial function is <code>main()</code>. For other
threads, the initial function is <code>operator()</code> of the function object
passed to the class <code>thread</code> constructor.</p>
<p>A thread of execution is said to be "finished" or "finished execution" when its
initial function returns or is terminated. This includes completion of all thread
cleanup handlers, and completion of the normal C++ function return behaviors, such
as destruction of automatic storage (stack) objects and releasing any associated
implementation resources.</p>
<p>A thread object has an associated state which is either "joinable" or
"non-joinable".</p>
<p>Except as described below, the policy used by an implementation of
<b>Boost.Threads</b> to schedule transitions between thread states is unspecified.</p>
<p><b>Note: </b>Just as the lifetime of a file may be different from the
lifetime of an iostream object which represents the file, the lifetime of a
thread of execution may be different from the <code>thread</code> object which
represents the thread of execution. In particular, after a call to <code>join()</code>,
the thread of execution will no longer exist even though the <code>thread</code>
object continues to exist until the end of its normal lifetime. The
converse is also possible; if a <code>thread</code> object is destroyed without
<code>join()</code> having first been called, the thread of execution continues until
its initial function completes.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <A href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</A>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost {
class thread : <a href="../../utility/utility.htm#noncopyable">boost::noncopyable</a> // Exposition only.
// Class thread meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
thread();
explicit thread(const boost::function0&lt;void&gt;&amp; threadfunc);
~thread();
bool operator==(const thread&amp; rhs) const;
bool operator!=(const thread&amp; rhs) const;
void join();
static void sleep(const xtime&amp; xt);
static void yield();
};
} // namespace boost
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructors</h3>
<pre>
thread();
</pre>
<p><b>Effects:</b> Constructs a <code>thread</code> object representing the current thread
of execution.</p>
<p><b>Postcondition:</b> <code>*this</code> is non-joinable.</p>
<p><b>Danger:</b> <code>*this</code> is valid only within the current thread.</p>
<pre>
thread(const <A href="../../function/index.html">boost::function0</A>&lt;void&gt;&amp; threadfunc);
</pre>
<p><b>Effects:</b> Starts a new thread of execution and constructs a <code>thread</code> object
representing it. Copies <code>threadfunc</code>
(which in turn copies the function object wrapped by <code>threadfunc</code>) to an
internal location which persists for the lifetime of the new thread of execution. Calls
<code>operator()</code> on the copy of the <code>threadfunc</code> function object in
the new thread of execution.</p>
<p><b>Postcondition:</b> <code>*this</code> is joinable.</p>
<p><b>Throws:</b> <code>boost::thread_resource_error</code> if a new thread of execution
cannot be started.</p>
<hr>
<h3>Destructor</h3>
<pre>
~thread();
</pre>
<p><b>Effects:</b> Destroys <code>*this</code>. The actual thread of execution may
continue to execute after the <code>thread</code> object has been destroyed.</p>
<p><b>Notes:</b> If <code>*this</code> is joinable the actual thread of execution
becomes "detached". Any resources used by the thread will be reclaimed when the
thread of execution completes. To ensure such a thread of execution runs to completion
before the <code>thread</code> object is destroyed, call <code>join()</code>.</p>
<hr>
<h3>Comparison Operators</h3>
<pre>
bool operator==(const thread&amp; rhs);
</pre>
<p><b>Requires:</b> The thread is non-terminated or <code>*this</code> is joinable.</p>
<p><b>Returns:</b> <code>true</code> if <code>*this</code> and <code>rhs</code>
represent the same thread of execution.</p>
<pre>
bool operator!=(const thread&amp; rhs);
</pre>
<p><b>Returns:</b> <code>!(*this==rhs)</code>.</p>
<hr>
<h3>join</h3>
<pre>
void join();
</pre>
<p><b>Requires:</b> <code>*this</code> is joinable.</p>
<p><b>Effects:</b> The current thread of execution blocks until the initial function of
the thread of execution represented by <code>*this</code> finishes and all resources
are reclaimed.</p>
<p><b>Postcondition:</b> <code>*this</code> is non-joinable.</p>
<p><b>Note:</b></p> If <code>*this == thread()</code> the result is implementation defined.
If the implementation doesn't detect this the result will be
<a href="definitions.html#Deadlock">deadlock</a>.</p>
<hr>
<h3>sleep</h3>
<pre>
static void sleep(const <a href="xtime.html">xtime</a>&amp; xt);
</pre>
<p><b>Effects:</b> The current thread of execution blocks until <code>xt</code> is
reached.</p>
<hr>
<h3>yield</h3>
<pre>
static void yield();
</pre>
<p><b>Effects:</b> The current thread of execution is placed in the "ready" state.</p>
<p><b>Notes:</b> Allow the current thread to give up the rest of its time slice
(or other scheduling quota) to another thread. Particularly useful in non-preemptive
implementations.</p>
<hr>
<h2><a name="Example">Example Usage</a></h2>
<pre>
#include &lt;boost/thread/thread.hpp&gt;
#include &lt;iostream&gt;
struct thread_alarm
{
thread_alarm(int secs) : m_secs(secs) { }
void operator()()
{
boost::xtime xt;
boost::xtime_get(&amp;xt, boost::TIME_UTC);
xt.sec += m_secs;
boost::thread::sleep(xt);
std::cout &lt;&lt; &quot;alarm sounded...&quot; &lt;&lt; std::endl;
}
int m_secs;
};
int main(int argc, char* argv[])
{
int secs = 5;
std::cout &lt;&lt; &quot;setting alarm for 5 seconds...&quot; &lt;&lt; std::endl;
boost::thread thrd(thread_alarm(secs));
thrd.join();
}
</pre>
<p>The output is:</p>
<pre>
setting alarm for 5 seconds...
alarm sounded...
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <A href="mailto:williamkempf@hotmail.com">William E. Kempf</A>
2001 all rights reserved.</i></p>
</body>
</html>

187
doc/thread_group.html Normal file
View File

@@ -0,0 +1,187 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, thread_group</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">thread_group</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#Synopsis">Synopsis</a><br>
<a href="#Members">Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <tt>thread_group</tt> class provides a container for easy grouping of threads to simplify several
common thread creation and management idioms.</p>
<p>All <tt>thread_group</tt> member functions are <a href="definitions.html#thread-safe">thread-safe</a>,
except destruction.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost
{
class thread_group : <a href="../../utility/utility.htm#noncopyable">boost::noncopyable</a>
{
public:
thread_group();
~thread_group();
thread* create_thread(const boost::function0&lt;void&gt;&amp; threadfunc);
void add_thread(thread* thrd);
void remove_thread(thread* thrd);
void join_all();
};
}
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
thread_group();
</pre>
<p><b>Effects:</b> Constructs an empty <tt>thread_group</tt> container.</p>
<hr>
<h3>Destructor</h3>
<pre>
~thread_group();
</pre>
<p><b>Effects:</b> Destroys each contained thread object.&nbsp; Destroys <code>*this</code>.</p>
<p><b>Notes:</b> Behavior is undefined if another thread references *this during
the execution of the destructor.</p>
<hr>
<h3>create_thread</h3>
<pre>
thread* create_thread(const boost::function0&lt;void&gt;&amp; threadfunc);
</pre>
<p><b>Effects:</b> Creates a new <tt>thread</tt> object that executes <tt>threadfunc</tt> and adds it to the
<tt>thread_group</tt> container object's list of managed <tt>thread</tt> objects.</p>
<p><b>Returns:</b> Pointer to the newly created thread.</p>
<hr>
<h3>add_thread</h3>
<pre>
void add_thread(thread* thrd);
</pre>
<p><b>Effects:</b> Adds <tt>thrd</tt> to the <tt>thread_group</tt> object's list of managed <tt>thread</tt>
objects. The <tt>thrd</tt> object must have been allocated via operator new and will
be deleted when the group is destroyed.</p>
<hr>
<h3>remove_thread</h3>
<pre>
void remove_thread(thread* thrd);
</pre>
<p><b>Effects:</b> Removes <code>*this</code>'s list of managed <tt>thread</tt>
objects.</p>
<p><b>Throws: </b>? if <tt>thrd</tt> is not it <code>*this</code>'s list of managed <tt>thread</tt>
objects.</p>
<hr>
<h3>join_all</h3>
<pre>
void join_all();
</pre>
<p><b>Effects:</b> Calls <code> join()</code> on each of the managed <tt>thread</tt> objects.</p>
<hr>
<h2><a name="Example">Example</a> Usage</h2>
<pre>
#include &lt;boost/thread/thread.hpp&gt;
#include &lt;iostream&gt;
int count = 0;
boost::mutex mutex;
void increment_count()
{
boost::mutex::lock lock(mutex);
std::cout &lt;&lt; &quot;count = &quot; &lt;&lt; ++count &lt;&lt; std::endl;
}
int main(int argc, char* argv[])
{
boost::thread_group threads;
for (int i = 0; i &lt; 10; ++i)
threads.create_thread(&amp;increment_count);
threads.join_all();
}
</pre>
<p>The output is:</p>
<pre>
count = 1
count = 2
count = 3
count = 4
count = 5
count = 6
count = 7
count = 8
count = 9
count = 10
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

View File

@@ -0,0 +1,80 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, BTL, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, thread_resource_error</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">thread_resource_error</h2>
</td>
</tr>
</table>
<hr>
<p><a href="#Introduction">Introduction</a><br>
<a href="#Header">Header</a><br>
<a href="#Synopsis">Synopsis</a><br>
<a href="#Members">Members</a><br>
<a href="#Example">Example</a></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <code>thread_resource_error</code> class defines an exception type that is thrown
by constructors in the <b>Boost.Threads</b> library when thread related resources
can not be
acquired. This does not include memory allocation failures which instead throw
std::bad_alloc.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/exceptions.hpp&gt;</a>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost
{
class thread_resource_error : public std::runtime_error
{
public:
thread_resource_error();
};
}
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
thread_resource_error();
</pre>
<p>Constructs a <code>thread_resource_error</code> object.</p>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

View File

@@ -0,0 +1,214 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, Boost.Threads, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, thread_specific_ptr</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">thread_specific_ptr</h2>
</td>
</tr>
</table>
<hr>
<p><A href="#Introduction">Introduction</A><br>
<A href="#Header">Header</A><br>
<A href="#Synopsis">Synopsis</A><br>
<A href="#Members">Members</A><br>
<A href="#Example">Example</A></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <code>thread_specific_ptr</code> class defines an interface for using thread
specific storage. Thread specific storage is data associated with individual threads
and is often used to make operations
<a href="definitions.html#Thread-safe">thread-safe</a> that rely on global data.</p>
<p>Template <code>thread_specific_ptr</code> stores a pointer to an object obtained via
<code>new</code> on a thread-by-thread basis and calls delete on the contained pointer
when the thread terminates. Each thread initially stores the null pointer in each
<code>thread_specific_ptr</code> instance.</p>
<p>The template <code>thread_specific_ptr</code> is useful in the following cases:</p>
<ul>
<li>An interface was original written assuming a single thread of control and is
being ported to a multi-threaded environment.</li>
<li>Each thread of control invokes sequences of methods that share data that must be
logically accessed through a globally visible access point, but are physically
unique for each thread, instead of being explicitly passed.</li>
</ul>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/tss.hpp">&lt;boost/thread/tss.hpp&gt;</a>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost {
template &lt;typename T&gt;
class thread_specific_ptr : private boost::noncopyable // Exposition only.
// Class thread_specific_ptr meets the <a href="overview.html#NonCopyable">NonCopyable</a> requirement.
{
public:
thread_specific_ptr();
~thread_specific_ptr();
T* get() const;
T* operator-&gt;() const;
T&amp; operator*() const;
T* release();
void reset(T* p=0);
};
} // namespace boost
</pre>
<h2><a name="Members">Members</a></h2>
<hr>
<h3>Constructor</h3>
<pre>
thread_specific_ptr();
</pre>
<p><b>Postconditions:</b> A thread specific storage has been reserved for use by *this
in all threads, with each thread initially storing a null pointer.</p>
<p><b>Requires:</b> The expression <code>delete get()</code> is well formed.</p>
<p><b>Throws:</b> <code>boost::thread_resource_error</code> if the necessary resources
can not be obtained.</p>
<p><b>Notes:</b> There is an implementation specific limit to the number of thread
specific storage objects that can be created, and this limit may be small.</p>
<hr>
<h3>Destructor</h3>
<pre>
~thread_specific_ptr();
</pre>
<p><b>Notes:</b> Does not destroy any data that may be stored in any thread's thread
specific storage. For this reason you should not destroy a
<code>thread_specific_ptr</code> object until you are certain there are no threads
running that have made use of its thread specific storage.</p>
<hr>
<h3>get</h3>
<pre>
T* get() const;
</pre>
<p><b>Returns:</b> The object stored in thread specific storage for the current thread
for *this.</p>
<p><b>Notes:</b> Each thread initially returns 0.</p>
<hr>
<h3>Smart Pointer Operations</h3>
<pre>
T* operator-&gt;() const;
</pre>
<p><b>Returns:</b> <code>get()</code></p>
<pre>
T&amp; operator*() const;
</pre>
<p><b>Returns:</b> <code>get()</code></p>
<p><b>Requires:</b> <code>get() != 0</code></p>
<hr>
<h3>Release</h3>
<pre>
T* release();
</pre>
<p><b>Returns:</b> <code>get()</code></p>
<p><b>Postcondition:</b> *this holds the null pointer for the current thread.</p>
<hr>
<h3>Reset</h3>
<pre>
void reset(T* p=0);
</pre>
<p><b>Effects:</b> If <code>get()!= p</code> then <code>delete get()</code>.</p>
<p><b>Postconditions:</b> <code>*this</code> holds the pointer <code>p</code> for
the current thread.</p>
<p><b>Notes:</b> The pointer will be deleted when the thread terminates.</p>
<hr>
<h2><a name="Example">Example Usage</a></h2>
<pre>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
#include <a href="../../../boost/thread/tss.hpp">&lt;boost/thread/tss.hpp&gt;</a>
#include &lt;cassert&gt;
boost::thread_specific_ptr&lt;int&gt; value;
void increment()
{
int* p = value.get();
++*p;
}
void thread_proc()
{
value.reset(new int(0)); // initialize the thread's storage
for (int i=0; i&lt;10; ++i)
{
increment();
int* p = value.get();
assert(*p == i+1);
}
}
int main(int argc, char* argv[])
{
boost::thread_group threads;
for (int i=0; i&lt;5; ++i)
threads.create_thread(&amp;thread_proc);
threads.join_all();
}
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

141
doc/xtime.html Normal file
View File

@@ -0,0 +1,141 @@
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<meta name="keywords" content="threads, Boost.Threads, thread library, C++">
<link rel="stylesheet" type="text/css" href="styles.css">
<title>Boost.Threads, xtime</title>
</head>
<body bgcolor="#FFFFFF" link="#0000FF" vlink="#800080">
<table border="0" cellpadding="7" cellspacing="0" width="100%">
<tr>
<td valign="top" width="300">
<h3><img src="../../../c++boost.gif" alt="C++ Boost" width="277" height="86"></h3>
</td>
<td valign="top">
<h1 align="center">Boost.Threads</h1>
<h2 align="center">xtime</h2>
</td>
</tr>
</table>
<hr>
<p><A href="#Introduction">Introduction</A><br>
<A href="#Header">Header</A><br>
<A href="#Synopsis">Synopsis</A><br>
<A href="#Reference">Reference</A><br>
<A href="#Example">Example</A></p>
<h2><a name="Introduction">Introduction</a></h2>
<p>The <code>xtime</code> type is used to represent a point on some time scale or
a duration in time. This type may be proposed for the C standard by Markus Kuhn.
<b>Boost.Threads</b> provides only a very minimal implementation of this proposal
and it's expected that a full implementation will be provided in Boost as a separate
library, at which time <b>Boost.Threads</b> will deprecate its implementation.</p>
<h2><a name="Header">Header</a></h2>
<pre>
#include <a href="../../../boost/thread/xtime.hpp">&lt;boost/thread/xtime.hpp&gt;</a>
</pre>
<h2><a name="Synopsis">Synopsis</a></h2>
<pre>
namespace boost {
enum
{
TIME_UTC=1,
};
struct xtime
{
#if defined(BOOST_NO_INT64_T)
int_fast32_t sec;
#else
int_fast64_t sec;
#endif
int_fast32_t nsec;
};
int xtime_get(struct xtime* xtp, int clock_type);
} // namespace boost
</pre>
<h2><a name="Reference">Reference</a></h2>
<hr>
<h3>TIME_UTC</h3>
<p>The clock type for Coordinated Universal Time (UTC). The epoch for this clock type
is 1970-01-01 00:00:00. This is the only clock type supported by <b>Boost.Threads</b>.</p>
<hr>
<h3>xtime</h3>
<pre>
struct xtime
{
#if defined(BOOST_NO_INT64_T)
int_fast32_t sec;
#else
int_fast64_t sec;
#endif
int_fast32_t nsec;
};
</pre>
<p><b>sec</b> represents the whole seconds that have passed since the epoch.</p>
<p><b>nsec</b> represents the nanoseconds since <code>sec.</code>
<hr>
<h3>xtime_get</h3>
<pre>
int xtime_get(struct xtime* xtp, int clock_type);
</pre>
<p><b>Postcondition:</b> <code>xtp</code> represents the current point in time
as a duration since the epoch specified by the <code>clock_type</code>.</p>
<p><b>Returns:</b> <code>clock_type</code> if successful, otherwise 0.
<p><b>Notes:</b> The resolution is implementation specific. For many
implementations the best resolution of time is far more than one nanosecond, and
even when the resolution is reasonably good, the latency of a call to <code>xtime_get()</code>
may be significant. For maximum portability, avoid durations of less than
one second.</p>
<hr>
<h2><a name="Example">Example Usage</a></h2>
<pre>
#include <a href="../../../boost/thread/thread.hpp">&lt;boost/thread/thread.hpp&gt;</a>
#include <a href="../../../boost/thread/tss.hpp">&lt;boost/thread/xtime.hpp&gt;</a>
int main(int argc, char* argv[])
{
boost::xtime xt;
boost::xtime_get(&amp;xt, boost::TIME_UTC);
xt.sec += 1;
boost::thread::sleep(xt); // Sleep for 1 second
}
</pre>
<hr>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %B, %Y" startspan -->01 October, 2001<!--webbot bot="Timestamp" endspan i-checksum="38291" -->
</p>
<p><i>© Copyright <a href="mailto:williamkempf@hotmail.com">William E. Kempf</a>
2001 all rights reserved.</i></p>
</body>
</html>

View File

105
example/monitor/monitor.cpp Normal file
View File

@@ -0,0 +1,105 @@
#include <vector>
#include <iostream>
#include <boost/thread/condition.hpp>
#include <boost/thread/mutex.hpp>
#include <boost/thread/recursive_mutex.hpp>
#include <boost/thread/thread.hpp>
namespace {
const int ITERS = 100;
boost::mutex io_mutex;
};
template <typename M>
class buffer_t
{
public:
typedef typename M::scoped_lock scoped_lock;
buffer_t(int n)
: p(0), c(0), full(0), buf(n)
{
}
void send(int m)
{
scoped_lock lk(mutex);
while (full == buf.size())
cond.wait(lk);
buf[p] = m;
p = (p+1) % buf.size();
++full;
cond.notify_all();
}
int receive()
{
scoped_lock lk(mutex);
while (full == 0)
cond.wait(lk);
int i = buf[c];
c = (c+1) % buf.size();
--full;
cond.notify_all();
return i;
}
static buffer_t& get_buffer()
{
static buffer_t buf(2);
return buf;
}
static void do_sender_thread()
{
for (int n = 0; n < ITERS; ++n)
{
get_buffer().send(n);
{
boost::mutex::scoped_lock lock(io_mutex);
std::cout << "sent: " << n << std::endl;
}
}
}
static void do_receiver_thread()
{
int n;
do
{
n = get_buffer().receive();
{
boost::mutex::scoped_lock lock(io_mutex);
std::cout << "received: " << n << std::endl;
}
} while (n < ITERS - 1);
}
private:
M mutex;
boost::condition cond;
unsigned int p, c, full;
std::vector<int> buf;
};
template <typename M>
void do_test(M* dummy=0)
{
typedef buffer_t<M> buffer_type;
buffer_type::get_buffer();
boost::thread thrd1(&buffer_type::do_sender_thread);
boost::thread thrd2(&buffer_type::do_receiver_thread);
thrd1.join();
thrd2.join();
}
void test_buffer()
{
do_test<boost::mutex>();
do_test<boost::recursive_mutex>();
}
int main()
{
test_buffer();
return 0;
}

View File

@@ -0,0 +1,171 @@
#include <boost/thread/mutex.hpp>
#include <boost/thread/condition.hpp>
#include <boost/thread/thread.hpp>
#include <boost/thread/xtime.hpp>
#include <iostream>
#include <time.h>
namespace
{
boost::mutex iomx;
};
class canteen
{
public:
canteen() : m_chickens(0) { }
void get(int id)
{
boost::mutex::scoped_lock lock(m_mutex);
while (m_chickens == 0)
{
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() << ") Phil" << id <<
": wot, no chickens? I'll WAIT ..." << std::endl;
}
m_condition.wait(lock);
}
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() << ") Phil" << id <<
": those chickens look good ... one please ..." << std::endl;
}
m_chickens--;
}
void put(int value)
{
boost::mutex::scoped_lock lock(m_mutex);
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() <<
") Chef: ouch ... make room ... this dish is very hot ..." << std::endl;
}
boost::xtime xt;
boost::xtime_get(&xt, boost::TIME_UTC);
xt.sec += 3;
boost::thread::sleep(xt);
m_chickens += value;
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() <<
") Chef: more chickens ... " << m_chickens <<
" now available ... NOTIFYING ..." << std::endl;
}
m_condition.notify_all();
}
private:
boost::mutex m_mutex;
boost::condition m_condition;
int m_chickens;
};
canteen g_canteen;
void chef()
{
const int chickens = 4;
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() << ") Chef: starting ..." << std::endl;
}
for (;;)
{
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() << ") Chef: cooking ..." << std::endl;
}
boost::xtime xt;
boost::xtime_get(&xt, boost::TIME_UTC);
xt.sec += 2;
boost::thread::sleep(xt);
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() << ") Chef: " << chickens
<< " chickens, ready-to-go ..." << std::endl;
}
g_canteen.put(chickens);
}
}
struct phil
{
phil(int id) : m_id(id) { }
void run() {
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() << ") Phil" << m_id << ": starting ..." << std::endl;
}
for (;;)
{
if (m_id > 0)
{
boost::xtime xt;
boost::xtime_get(&xt, boost::TIME_UTC);
xt.sec += 3;
boost::thread::sleep(xt);
}
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() << ") Phil" << m_id
<< ": gotta eat ..." << std::endl;
}
g_canteen.get(m_id);
{
boost::mutex::scoped_lock lock(iomx);
std::cout << "(" << clock() << ") Phil" << m_id
<< ": mmm ... that's good ..." << std::endl;
}
}
}
static void do_thread(void* param) {
static_cast<phil*>(param)->run();
}
int m_id;
};
struct thread_adapt
{
thread_adapt(void (*func)(void*), void* param) : _func(func), _param(param) { }
int operator()() const
{
_func(_param);
return 0;
}
void (*_func)(void*);
void* _param;
};
class thread_adapter
{
public:
thread_adapter(void (*func)(void*), void* param) : _func(func), _param(param) { }
void operator()() const { _func(_param); }
private:
void (*_func)(void*);
void* _param;
};
int main(int argc, char* argv[])
{
boost::thread thrd_chef(&chef);
phil p[] = { phil(0), phil(1), phil(2), phil(3), phil(4) };
boost::thread thrd_phil0(thread_adapter(&phil::do_thread, &p[0]));
boost::thread thrd_phil1(thread_adapter(&phil::do_thread, &p[1]));
boost::thread thrd_phil2(thread_adapter(&phil::do_thread, &p[2]));
boost::thread thrd_phil3(thread_adapter(&phil::do_thread, &p[3]));
boost::thread thrd_phil4(thread_adapter(&phil::do_thread, &p[4]));
thrd_chef.join();
thrd_phil0.join();
thrd_phil1.join();
thrd_phil2.join();
thrd_phil3.join();
thrd_phil4.join();
return 0;
}

121
example/tennis/tennis.cpp Normal file
View File

@@ -0,0 +1,121 @@
#include <boost/thread/mutex.hpp>
#include <boost/thread/condition.hpp>
#include <boost/thread/semaphore.hpp>
#include <boost/thread/thread.hpp>
#include <boost/thread/xtime.hpp>
#include <iostream>
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
# include <process.h>
#endif
enum game_state
{
START,
PLAYER_A,
PLAYER_B,
GAME_OVER,
ONE_PLAYER_GONE,
BOTH_PLAYERS_GONE
};
int state;
boost::mutex mutex;
boost::condition cond;
char* player_name(int state)
{
if (state == PLAYER_A)
return "PLAYER-A";
if (state == PLAYER_B)
return "PLAYER-B";
throw "bad player";
return 0;
}
void player(void* param)
{
boost::mutex::scoped_lock lock(mutex);
int active = (int)param;
int other = active == PLAYER_A ? PLAYER_B : PLAYER_A;
while (state < GAME_OVER)
{
std::cout << player_name(active) << ": Play." << std::endl;
state = other;
cond.notify_all();
do
{
cond.wait(lock);
if (state == other)
std::cout << "---" << player_name(active) << ": Spurious wakeup!" << std::endl;
} while (state == other);
}
++state;
std::cout << player_name(active) << ": Gone." << std::endl;
cond.notify_all();
}
struct thread_adapt
{
thread_adapt(void (*func)(void*), void* param) : _func(func), _param(param) { }
int operator()() const
{
_func(_param);
return 0;
}
void (*_func)(void*);
void* _param;
};
class thread_adapter
{
public:
thread_adapter(void (*func)(void*), void* param) : _func(func), _param(param) { }
void operator()() const { _func(_param); }
private:
void (*_func)(void*);
void* _param;
};
int main(int argc, char* argv[])
{
state = START;
boost::thread thrda(thread_adapter(&player, (void*)PLAYER_A));
boost::thread thrdb(thread_adapter(&player, (void*)PLAYER_B));
boost::xtime xt;
boost::xtime_get(&xt, boost::TIME_UTC);
xt.sec += 1;
boost::thread::sleep(xt);
{
boost::mutex::scoped_lock lock(mutex);
std::cout << "---Noise ON..." << std::endl;
}
for (int i = 0; i < 1000000; ++i)
cond.notify_all();
{
boost::mutex::scoped_lock lock(mutex);
std::cout << "---Noise OFF..." << std::endl;
state = GAME_OVER;
cond.notify_all();
do
{
cond.wait(lock);
} while (state != BOTH_PLAYERS_GONE);
}
std::cout << "GAME OVER" << std::endl;
thrda.join();
thrdb.join();
return 0;
}

View File

@@ -0,0 +1,156 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_CONDITION_WEK070601_HPP
#define BOOST_CONDITION_WEK070601_HPP
#include <boost/config.hpp>
#ifndef BOOST_HAS_THREADS
# error Thread support is unavailable!
#endif
#include <boost/thread/exceptions.hpp>
#include <boost/utility.hpp>
#if defined(BOOST_HAS_PTHREADS)
# include <pthread.h>
#endif
namespace boost {
struct xtime;
class condition : private noncopyable
{
public:
condition();
~condition();
void notify_one();
void notify_all();
template <typename L>
void wait(L& lock)
{
if (!lock)
throw lock_error();
do_wait(lock.m_mutex);
}
template <typename L, typename Pr>
void wait(L& lock, Pr pred)
{
if (!lock)
throw lock_error();
while (!pred())
do_wait(lock.m_mutex);
}
template <typename L>
bool timed_wait(L& lock, const xtime& xt)
{
if (!lock)
throw lock_error();
return do_timed_wait(lock.m_mutex, xt);
}
template <typename L, typename Pr>
bool timed_wait(L& lock, const xtime& xt, Pr pred)
{
if (!lock)
throw lock_error();
while (!pred())
{
if (!do_timed_wait(lock.m_mutex, xt))
return false;
}
return true;
}
private:
template <typename M>
void do_wait(M& mutex)
{
#if defined(BOOST_HAS_WINTHREADS)
enter_wait();
#endif
typename M::cv_state state;
mutex.do_unlock(state);
#if defined(BOOST_HAS_PTHREADS)
do_wait(state.pmutex);
#elif defined(BOOST_HAS_WINTHREADS)
do_wait();
#endif
mutex.do_lock(state);
}
template <typename M>
bool do_timed_wait(M& mutex, const xtime& xt)
{
#if defined(BOOST_HAS_WINTHREADS)
enter_wait();
#endif
typename M::cv_state state;
mutex.do_unlock(state);
bool ret = false;
#if defined(BOOST_HAS_PTHREADS)
ret = do_timed_wait(xt, state.pmutex);
#elif defined(BOOST_HAS_WINTHREADS)
ret = do_timed_wait(xt);
#endif
mutex.do_lock(state);
return ret;
}
#if defined(BOOST_HAS_WINTHREADS)
void enter_wait();
void do_wait();
bool do_timed_wait(const xtime& xt);
#elif defined(BOOST_HAS_PTHREADS)
void do_wait(pthread_mutex_t* pmutex);
bool do_timed_wait(const xtime& xt, pthread_mutex_t* pmutex);
#endif
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_gate;
unsigned long m_queue;
unsigned long m_mutex;
unsigned m_gone; // # threads that timed out and never made it to the m_queue
unsigned long m_blocked; // # threads m_blocked m_waiting for the condition
unsigned m_waiting; // # threads m_waiting no longer m_waiting for the condition but still
// m_waiting to be removed from the m_queue
#elif defined(BOOST_HAS_PTHREADS)
pthread_cond_t m_condition;
#endif
};
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 22 May 01 WEKEMPF Modified to use xtime for time outs.
// 23 May 01 WEKEMPF Removed "duration" timed_waits, as they are too difficult
// to use with spurious wakeups.
#endif // BOOST_CONDITION_WEK070601_HPP

View File

@@ -0,0 +1,100 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
// This file is used to configure Boost.Threads during development
// in order to decouple dependency on any Boost release. Once
// accepted into Boost these contents will be moved to <boost/config>
// or some other appropriate build configuration and all
// #include <boost/thread/config.hpp> statements will be changed
// accordingly.
#ifndef BOOST_THREAD_CONFIG_WEK070601_HPP
#define BOOST_THREAD_CONFIG_WEK070601_HPP
#include <boost/config.hpp>
#error "Included <boost/thread/config.hpp>"
/*// Define if threading support is enabled for the toolset.
#undef BOOST_HAS_THREADS
// Define if threading should be implemented in terms of Win32 threads.
#undef BOOST_HAS_WINTHREADS
// Define if threading should be implemented in terms of POSIX threads.
#undef BOOST_HAS_PTHREADS
// Define if BOOST_HAS_PTHREADS and pthread_delay_np() exists.
#undef BOOST_HAS_PTHREAD_DELAY_NP
// Define if BOOST_HAS_PTHREADS and not BOOST_HAS_PTHREAD_DELAY_NP
// but nanosleep can be used instead.
#undef BOOST_HAS_NANOSLEEP
// Define if BOOST_HAS_PTHREADS and pthread_yield() exists.
#undef BOOST_HAS_PTHREAD_YIELD
// Define if BOOST_HAS_PTHREADS and not BOOST_HAS_PTHREAD_YIELD and
// sched_yield() exists.
#undef BOOST_HAS_SCHED_YIELD
// Define if gettimeofday() exists.
#undef BOOST_HAS_GETTIMEOFDAY
// Define if not BOOST_HAS_GETTIMEOFDAY and clock_gettime() exists.
#undef BOOST_HAS_CLOCK_GETTIME
// Define if not BOOST_HAS_GETTIMEOFDAY and not BOOST_HAS_CLOCK_GETTIME and
// GetSystemTimeAsFileTime() can be called with an FTIME structure.
#undef BOOST_HAS_FTIME
// Define if pthread_mutexattr_settype and pthread_mutexattr_gettype exist.
#undef BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE
// Here we'll set up known compiler options.
#if defined(BOOST_MSVC)
# if defined(_MT)
# define BOOST_HAS_THREADS
# endif
# define BOOST_HAS_WINTHREADS // comment out this to test pthreads-win32.
# if !defined(BOOST_HAS_WINTHREADS)
# define BOOST_HAS_PTHREADS
# define BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE
# define PtW32NoCatchWarn
# pragma comment(lib, "pthreadVCE.lib")
# endif
# define BOOST_HAS_FTIME
// pdm: this is for linux - is there a better #define to #if on?
// wek: not sure how else to do this, but GNU CC on Win32 should probably
// use BOOST_HAS_WINTHREADS, and I expect there will be other
// platform specific variations for this compiler toolset. Need
// to decide how to handle this.
#elif defined( __GNUC__ )
# define BOOST_HAS_THREADS
# define BOOST_HAS_PTHREADS
# define BOOST_HAS_NANOSLEEP
# define BOOST_HAS_GETTIMEOFDAY
// pdm: From the pthread.h header, one of these macros
// must be defined for this stuff to exist.
// wek: This seems like a harmless enough method to determine these
// switches, but one should note that some implementations may not
// use these. Notably, pthreads-win32 doesn't define either
// __USE_UNIX98 or __USE_GNU.
# if defined( __USE_UNIX98 )
# define BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE
# elif defined( __USE_GNU )
# define BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE
# define BOOST_HAS_PTHREAD_YIELD
# endif
#endif*/
#endif // BOOST_THREAD_CONFIG_WEK070601_HPP

View File

@@ -0,0 +1,172 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_XLOCK_WEK070601_HPP
#define BOOST_XLOCK_WEK070601_HPP
#include <boost/utility.hpp>
#include <boost/thread/exceptions.hpp>
namespace boost {
class condition;
struct xtime;
namespace detail { namespace thread {
template <typename Mutex>
class scoped_lock : private noncopyable
{
public:
typedef Mutex mutex_type;
explicit scoped_lock(Mutex& mx, bool initially_locked=true)
: m_mutex(mx), m_locked(false)
{
if (initially_locked) lock();
}
~scoped_lock()
{
if (m_locked) unlock();
}
void lock()
{
if (m_locked) throw lock_error();
m_mutex.do_lock();
m_locked = true;
}
void unlock()
{
if (!m_locked) throw lock_error();
m_mutex.do_unlock();
m_locked = false;
}
bool locked() const { return m_locked; }
operator const void*() const { return m_locked ? this : 0; }
private:
friend class boost::condition;
Mutex& m_mutex;
bool m_locked;
};
template <typename TryMutex>
class scoped_try_lock : private noncopyable
{
public:
typedef TryMutex mutex_type;
explicit scoped_try_lock(TryMutex& mx)
: m_mutex(mx), m_locked(false)
{
try_lock();
}
scoped_try_lock(TryMutex& mx, bool initially_locked)
: m_mutex(mx), m_locked(false)
{
if (initially_locked) lock();
}
~scoped_try_lock()
{
if (m_locked) unlock();
}
void lock()
{
if (m_locked) throw lock_error();
m_mutex.do_lock();
m_locked = true;
}
bool try_lock()
{
if (m_locked) throw lock_error();
return (m_locked = m_mutex.do_trylock());
}
void unlock()
{
if (!m_locked) throw lock_error();
m_mutex.do_unlock();
m_locked = false;
}
bool locked() const { return m_locked; }
operator const void*() const { return m_locked ? this : 0; }
private:
friend class boost::condition;
TryMutex& m_mutex;
bool m_locked;
};
template <typename TimedMutex>
class scoped_timed_lock : private noncopyable
{
public:
typedef TimedMutex mutex_type;
scoped_timed_lock(TimedMutex& mx, const xtime& xt)
: m_mutex(mx), m_locked(false)
{
timed_lock(xt);
}
scoped_timed_lock(TimedMutex& mx, bool initially_locked)
: m_mutex(mx), m_locked(false)
{
if (initially_locked) lock();
}
~scoped_timed_lock()
{
if (m_locked) unlock();
}
void lock()
{
if (m_locked) throw lock_error();
m_mutex.do_lock();
m_locked = true;
}
bool timed_lock(const xtime& xt)
{
if (m_locked) throw lock_error();
return (m_locked = m_mutex.do_timedlock(xt));
}
void unlock()
{
if (!m_locked) throw lock_error();
m_mutex.do_unlock();
m_locked = false;
}
bool locked() const { return m_locked; }
operator const void*() const { return m_locked ? this : 0; }
private:
friend class boost::condition;
TimedMutex& m_mutex;
bool m_locked;
};
} // namespace thread
} // namespace detail
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 22 May 01 WEKEMPF Modified to use xtime for time outs.
// 30 Jul 01 WEKEMPF Moved lock types into boost::detail::thread. Renamed some types.
// Added locked() methods.
#endif // BOOST_XLOCK_WEK070601_HPP

View File

@@ -0,0 +1,46 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
// This file is used to configure Boost.Threads during development
// in order to decouple dependency on any Boost release. Once
// accepted into Boost these contents will be moved to <boost/config>
// or some other appropriate build configuration and all
// #include <boost/thread/config.hpp> statements will be changed
// accordingly.
#ifndef BOOST_THREAD_EXCEPTIONS_PDM070801_H
#define BOOST_THREAD_EXCEPTIONS_PDM070801_H
// pdm: Sorry, but this class is used all over the place & I end up
// with recursive headers if I don't separate it
// wek: Not sure why recursive headers would cause compilation problems
// given the include guards, but regardless it makes sense to
// seperate this out any way.
#include <stdexcept>
namespace boost {
class lock_error : public std::runtime_error
{
public:
lock_error() : std::runtime_error("thread lock error") { }
};
class thread_resource_error : public std::runtime_error
{
public:
thread_resource_error() : std::runtime_error("thread resource error") { }
};
} // namespace boost
#endif // BOOST_THREAD_CONFIG_PDM070801_H

View File

@@ -0,0 +1,146 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_MUTEX_WEK070601_HPP
#define BOOST_MUTEX_WEK070601_HPP
#include <boost/config.hpp>
#ifndef BOOST_HAS_THREADS
# error Thread support is unavailable!
#endif
#include <boost/utility.hpp>
#include <boost/thread/detail/lock.hpp>
#if defined(BOOST_HAS_PTHREADS)
# include <pthread.h>
#endif
namespace boost {
class condition;
struct xtime;
class mutex : private noncopyable
{
public:
friend class detail::thread::scoped_lock<mutex>;
friend class condition;
typedef detail::thread::scoped_lock<mutex> scoped_lock;
mutex();
~mutex();
private:
#if defined(BOOST_HAS_WINTHREADS)
typedef void* cv_state;
#elif defined(BOOST_HAS_PTHREADS)
struct cv_state
{
pthread_mutex_t* pmutex;
};
#endif
void do_lock();
void do_unlock();
void do_lock(cv_state& state);
void do_unlock(cv_state& state);
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_mutex;
#elif defined(BOOST_HAS_PTHREADS)
pthread_mutex_t m_mutex;
#endif
};
class try_mutex : private noncopyable
{
public:
friend class detail::thread::scoped_lock<try_mutex>;
friend class detail::thread::scoped_try_lock<try_mutex>;
friend class condition;
typedef detail::thread::scoped_lock<try_mutex> scoped_lock;
typedef detail::thread::scoped_try_lock<try_mutex> scoped_try_lock;
try_mutex();
~try_mutex();
private:
#if defined(BOOST_HAS_WINTHREADS)
typedef void* cv_state;
#elif defined(BOOST_HAS_PTHREADS)
struct cv_state
{
pthread_mutex_t* pmutex;
};
#endif
void do_lock();
bool do_trylock();
void do_unlock();
void do_lock(cv_state& state);
void do_unlock(cv_state& state);
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_mutex;
#elif defined(BOOST_HAS_PTHREADS)
pthread_mutex_t m_mutex;
#endif
};
class timed_mutex : private noncopyable
{
public:
friend class detail::thread::scoped_lock<timed_mutex>;
friend class detail::thread::scoped_try_lock<timed_mutex>;
friend class detail::thread::scoped_timed_lock<timed_mutex>;
friend class condition;
typedef detail::thread::scoped_lock<timed_mutex> scoped_lock;
typedef detail::thread::scoped_try_lock<timed_mutex> scoped_try_lock;
typedef detail::thread::scoped_timed_lock<timed_mutex> scoped_timed_lock;
timed_mutex();
~timed_mutex();
private:
#if defined(BOOST_HAS_WINTHREADS)
typedef void* cv_state;
#elif defined(BOOST_HAS_PTHREADS)
struct cv_state
{
pthread_mutex_t* pmutex;
};
#endif
void do_lock();
bool do_trylock();
bool do_timedlock(const xtime& xt);
void do_unlock();
void do_lock(cv_state& state);
void do_unlock(cv_state& state);
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_mutex;
#elif defined(BOOST_HAS_PTHREADS)
pthread_mutex_t m_mutex;
pthread_cond_t m_condition;
bool m_locked;
#endif
};
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 22 May 01 WEKEMPF Modified to use xtime for time outs. Factored out
// to three classes, mutex, try_mutex and timed_mutex.
#endif // BOOST_MUTEX_WEK070601_HPP

View File

@@ -0,0 +1,45 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_ONCE_WEK080101_HPP
#define BOOST_ONCE_WEK080101_HPP
#include <boost/config.hpp>
#ifndef BOOST_HAS_THREADS
# error Thread support is unavailable!
#endif
#if defined(BOOST_HAS_PTHREADS)
# include <pthread.h>
#endif
namespace boost {
#if defined(BOOST_HAS_PTHREADS)
typedef pthread_once_t once_flag;
const once_flag once_init = PTHREAD_ONCE_INIT;
#elif defined(BOOST_HAS_WINTHREADS)
typedef bool once_flag;
const once_flag once_init = false;
#endif
void call_once(void (*func)(), once_flag& flag);
} // namespace boost
// Change Log:
// 1 Aug 01 WEKEMPF Initial version.
#endif // BOOST_ONCE_WEK080101_HPP

View File

@@ -0,0 +1,167 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_RECURSIVE_MUTEX_WEK070601_HPP
#define BOOST_RECURSIVE_MUTEX_WEK070601_HPP
#include <boost/config.hpp>
#ifndef BOOST_HAS_THREADS
# error Thread support is unavailable!
#endif
#include <boost/utility.hpp>
#include <boost/thread/detail/lock.hpp>
#if defined(BOOST_HAS_PTHREADS)
# include <pthread.h>
#endif
namespace boost {
class condition;
struct xtime;
class recursive_mutex : private noncopyable
{
public:
friend class detail::thread::scoped_lock<recursive_mutex>;
friend class condition;
typedef detail::thread::scoped_lock<recursive_mutex> scoped_lock;
recursive_mutex();
~recursive_mutex();
private:
#if defined(BOOST_HAS_WINTHREADS)
typedef std::size_t cv_state;
#elif defined(BOOST_HAS_PTHREADS)
struct cv_state
{
long count;
pthread_mutex_t* pmutex;
};
#endif
void do_lock();
void do_unlock();
void do_lock(cv_state& state);
void do_unlock(cv_state& state);
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_mutex;
unsigned long m_count;
#elif defined(BOOST_HAS_PTHREADS)
pthread_mutex_t m_mutex;
unsigned m_count;
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
pthread_cond_t m_unlocked;
pthread_t m_thread_id;
bool m_valid_id;
# endif
#endif
};
class recursive_try_mutex : private noncopyable
{
public:
friend class detail::thread::scoped_lock<recursive_try_mutex>;
friend class detail::thread::scoped_try_lock<recursive_try_mutex>;
friend class condition;
typedef detail::thread::scoped_lock<recursive_try_mutex> scoped_lock;
typedef detail::thread::scoped_try_lock<recursive_try_mutex> scoped_try_lock;
recursive_try_mutex();
~recursive_try_mutex();
private:
#if defined(BOOST_HAS_WINTHREADS)
typedef std::size_t cv_state;
#elif defined(BOOST_HAS_PTHREADS)
struct cv_state
{
long count;
pthread_mutex_t* pmutex;
};
#endif
void do_lock();
bool do_trylock();
void do_unlock();
void do_lock(cv_state& state);
void do_unlock(cv_state& state);
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_mutex;
unsigned long m_count;
#elif defined(BOOST_HAS_PTHREADS)
pthread_mutex_t m_mutex;
unsigned m_count;
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
pthread_cond_t m_unlocked;
pthread_t m_thread_id;
bool m_valid_id;
# endif
#endif
};
class recursive_timed_mutex : private noncopyable
{
public:
friend class detail::thread::scoped_lock<recursive_timed_mutex>;
friend class detail::thread::scoped_try_lock<recursive_timed_mutex>;
friend class detail::thread::scoped_timed_lock<recursive_timed_mutex>;
friend class condition;
typedef detail::thread::scoped_lock<recursive_timed_mutex> scoped_lock;
typedef detail::thread::scoped_try_lock<recursive_timed_mutex> scoped_try_lock;
typedef detail::thread::scoped_timed_lock<recursive_timed_mutex> scoped_timed_lock;
recursive_timed_mutex();
~recursive_timed_mutex();
private:
#if defined(BOOST_HAS_WINTHREADS)
typedef std::size_t cv_state;
#elif defined(BOOST_HAS_PTHREADS)
struct cv_state
{
long count;
pthread_mutex_t* pmutex;
};
#endif
void do_lock();
bool do_trylock();
bool do_timedlock(const xtime& xt);
void do_unlock();
void do_lock(cv_state& state);
void do_unlock(cv_state& state);
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_mutex;
unsigned long m_count;
#elif defined(BOOST_HAS_PTHREADS)
pthread_mutex_t m_mutex;
pthread_cond_t m_unlocked;
pthread_t m_thread_id;
bool m_valid_id;
unsigned m_count;
#endif
};
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 1 Jun 01 WEKEMPF Modified to use xtime for time outs. Factored out
// to three classes, mutex, try_mutex and timed_mutex.
// 11 Jun 01 WEKEMPF Modified to use PTHREAD_MUTEX_RECURSIVE if available.
#endif // BOOST_RECURSIVE_MUTEX_WEK070601_HPP

View File

@@ -0,0 +1,57 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_SEMAPHORE_WEK070601_HPP
#define BOOST_SEMAPHORE_WEK070601_HPP
#include <boost/config.hpp>
#ifndef BOOST_HAS_THREADS
# error Thread support is unavailable!
#endif
#include <boost/utility.hpp>
#if defined(BOOST_HAS_PTHREADS)
# include <pthread.h>
#endif
namespace boost {
struct xtime;
class semaphore : private noncopyable
{
public:
explicit semaphore(unsigned count=0, unsigned max=0);
~semaphore();
bool up(unsigned count=1, unsigned* prev=0);
void down();
bool down(const xtime& xt);
private:
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_sema;
#elif defined(BOOST_HAS_PTHREADS)
pthread_mutex_t m_mutex;
pthread_cond_t m_condition;
unsigned m_available;
unsigned m_max;
#endif
};
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 22 May 01 WEKEMPF Modified to use xtime for time outs.
#endif // BOOST_SEMAPHORE_WEK070601_HPP

View File

@@ -0,0 +1,84 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_THREAD_WEK070601_HPP
#define BOOST_THREAD_WEK070601_HPP
#include <boost/config.hpp>
#ifndef BOOST_HAS_THREADS
# error Thread support is unavailable!
#endif
#include <boost/utility.hpp>
#include <boost/function.hpp>
#include <boost/thread/mutex.hpp>
#include <list>
#include <memory>
#if defined(BOOST_HAS_PTHREADS)
# include <pthread.h>
# include <boost/thread/condition.hpp>
#endif
namespace boost {
struct xtime;
class thread : private noncopyable
{
public:
thread();
explicit thread(const function0<void>& threadfunc);
~thread();
bool operator==(const thread& other) const;
bool operator!=(const thread& other) const;
void join();
static void sleep(const xtime& xt);
static void yield();
private:
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_thread;
unsigned int m_id;
#elif defined(BOOST_HAS_PTHREADS)
private:
pthread_t m_thread;
#endif
bool m_joinable;
};
class thread_group : private noncopyable
{
public:
thread_group();
~thread_group();
thread* create_thread(const function0<void>& threadfunc);
void add_thread(thread* thrd);
void remove_thread(thread* thrd);
void join_all();
private:
std::list<thread*> m_threads;
mutex m_mutex;
};
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 1 Jun 01 WEKEMPF Added boost::thread initial implementation.
// 3 Jul 01 WEKEMPF Redesigned boost::thread to be noncopyable.
#endif // BOOST_THREAD_WEK070601_HPP

View File

@@ -0,0 +1,72 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_TSS_WEK070601_HPP
#define BOOST_TSS_WEK070601_HPP
#include <boost/config.hpp>
#ifndef BOOST_HAS_THREADS
# error Thread support is unavailable!
#endif
#include <boost/utility.hpp>
#if defined(BOOST_HAS_PTHREADS)
# include <pthread.h>
#endif
namespace boost {
namespace detail {
class tss : private noncopyable
{
public:
tss(void (*cleanup)(void*)=0);
~tss();
void* get() const;
bool set(void* value);
private:
#if defined(BOOST_HAS_WINTHREADS)
unsigned long m_key;
void (*m_cleanup)(void*);
#elif defined(BOOST_HAS_PTHREADS)
pthread_key_t m_key;
#endif
};
}
template <typename T>
class thread_specific_ptr : private noncopyable
{
public:
thread_specific_ptr() : m_tss(&thread_specific_ptr<T>::cleanup) { }
T* get() const { return static_cast<T*>(m_tss.get()); }
T* operator->() const { return get(); }
T& operator*() const { return *get(); }
T* release() { T* temp = get(); m_tss.set(0); return temp; }
void reset(T* p=0) { T* cur = get(); if (cur == p) return; delete cur; m_tss.set(p); }
private:
static void cleanup(void* p) { delete static_cast<T*>(p); }
mutable detail::tss m_tss;
};
} // namespace boost
// Change Log:
// 6 Jun 01 WEKEMPF Initial version.
#endif // BOOST_TSS_WEK070601_HPP

View File

@@ -0,0 +1,50 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#ifndef BOOST_XTIME_WEK070601_HPP
#define BOOST_XTIME_WEK070601_HPP
#include <boost/cstdint.hpp>
#include <boost/config.hpp>
namespace boost {
enum
{
TIME_UTC=1,
TIME_TAI,
TIME_MONOTONIC,
TIME_PROCESS,
TIME_THREAD,
TIME_LOCAL,
TIME_SYNC,
TIME_RESOLUTION
};
struct xtime
{
#if defined(BOOST_NO_INT64_T)
int_fast32_t sec;
#else
int_fast64_t sec;
#endif
int_fast32_t nsec;
};
int xtime_get(struct xtime* xtp, int clock_type);
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
#endif // BOOST_XTIME_WEK070601_HPP

80
src/_atomic.cpp Normal file
View File

@@ -0,0 +1,80 @@
/*
*
* Copyright (C) 2001
* William E. Kempf
*
* Permission to use, copy, modify, distribute and sell this software
* and its documentation for any purpose is hereby granted without fee,
* provided that the above copyright notice appear in all copies and
* that both that copyright notice and this permission notice appear
* in supporting documentation. William E. Kempf makes no representations
* about the suitability of this software for any purpose.
* It is provided "as is" without express or implied warranty.
*
* Revision History (excluding minor changes for specific compilers)
* 8 Feb 01 Initial version.
*/
#include <boost/thread/atomic.hpp>
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
#endif
namespace boost {
atomic_t::value_type read(const atomic_t& x)
{
return x._value;
}
#if defined(BOOST_HAS_WINTHREADS)
atomic_t::value_type increment(atomic_t& x)
{
return InterlockedIncrement(const_cast<long*>(&x._value));
}
atomic_t::value_type decrement(atomic_t& x)
{
return InterlockedDecrement(const_cast<long*>(&x._value));
}
atomic_t::value_type swap(atomic_t& x, atomic_t::value_type y)
{
return InterlockedExchange(const_cast<long*>(&x._value), y);
}
atomic_t::value_type compare_swap(atomic_t& x, atomic_t::value_type y, atomic_t::value_type z)
{
return InterlockedCompareExchange(const_cast<long*>(&x._value), y, z);
}
#else
atomic_t::value_type increment(atomic_t& x)
{
mutex::lock lock(x._mutex);
return ++x._value;
}
atomic_t::value_type decrement(atomic_t& x)
{
mutex::lock lock(x._mutex);
return --x._value;
}
atomic_t::value_type swap(atomic_t& x, atomic_t::value_type y)
{
mutex::lock lock(x._mutex);
atomic_t::value_type temp = x._value;
x._value = y;
return temp;
}
atomic_t::value_type compare_swap(atomic_t& x, atomic_t::value_type y, atomic_t::value_type z)
{
mutex::lock lock(x._mutex);
atomic_t::value_type temp = x._value;
if (temp == z)
x._value = y;
return temp;
}
#endif
} // namespace boost

360
src/condition.cpp Normal file
View File

@@ -0,0 +1,360 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#include <boost/thread/condition.hpp>
#include <boost/thread/xtime.hpp>
#include <boost/thread/thread.hpp>
#include <boost/thread/exceptions.hpp>
#include <boost/limits.hpp>
#include <cassert>
#include "timeconv.inl"
#if defined(BOOST_HAS_WINTHREADS)
# define NOMINMAX
# include <windows.h>
#elif defined(BOOST_HAS_PTHREADS)
# include <errno.h>
#endif
namespace boost {
#if defined(BOOST_HAS_WINTHREADS)
condition::condition()
: m_gone(0), m_blocked(0), m_waiting(0)
{
m_gate = reinterpret_cast<unsigned long>(CreateSemaphore(0, 1, 1, 0));
m_queue = reinterpret_cast<unsigned long>(CreateSemaphore(0, 0, std::numeric_limits<long>::max(), 0));
m_mutex = reinterpret_cast<unsigned long>(CreateMutex(0, 0, 0));
if (!m_gate || !m_queue || !m_mutex)
{
int res = 0;
if (m_gate)
{
res = CloseHandle(reinterpret_cast<HANDLE>(m_gate));
assert(res);
}
if (m_queue)
{
res = CloseHandle(reinterpret_cast<HANDLE>(m_queue));
assert(res);
}
if (m_mutex)
{
res = CloseHandle(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
throw thread_resource_error();
}
}
condition::~condition()
{
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_gate));
assert(res);
res = CloseHandle(reinterpret_cast<HANDLE>(m_queue));
assert(res);
res = CloseHandle(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void condition::notify_one()
{
unsigned signals = 0;
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
if (m_waiting != 0) // the m_gate is already closed
{
if (m_blocked == 0)
{
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
return;
}
++m_waiting;
--m_blocked;
}
else
{
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_gate), INFINITE);
assert(res == WAIT_OBJECT_0);
if (m_blocked > m_gone)
{
if (m_gone != 0)
{
m_blocked -= m_gone;
m_gone = 0;
}
signals = m_waiting = 1;
--m_blocked;
}
else
{
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0);
assert(res);
}
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
if (signals)
{
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_queue), signals, 0);
assert(res);
}
}
}
void condition::notify_all()
{
unsigned signals = 0;
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
if (m_waiting != 0) // the m_gate is already closed
{
if (m_blocked == 0)
{
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
return;
}
m_waiting += (signals = m_blocked);
m_blocked = 0;
}
else
{
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_gate), INFINITE);
assert(res == WAIT_OBJECT_0);
if (m_blocked > m_gone)
{
if (m_gone != 0)
{
m_blocked -= m_gone;
m_gone = 0;
}
signals = m_waiting = m_blocked;
m_blocked = 0;
}
else
{
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0);
assert(res);
}
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
if (signals)
{
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_queue), signals, 0);
assert(res);
}
}
}
void condition::enter_wait()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_gate), INFINITE);
assert(res == WAIT_OBJECT_0);
++m_blocked;
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0);
assert(res);
}
void condition::do_wait()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_queue), INFINITE);
assert(res == WAIT_OBJECT_0);
unsigned was_waiting=0;
unsigned was_gone=0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
was_waiting = m_waiting;
was_gone = m_gone;
if (was_waiting != 0)
{
if (--m_waiting == 0)
{
if (m_blocked != 0)
{
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0); // open m_gate
assert(res);
was_waiting = 0;
}
else if (m_gone != 0)
m_gone = 0;
}
}
else if (++m_gone == (std::numeric_limits<unsigned>::max() / 2))
{
// timeout occured, normalize the m_gone count
// this may occur if many calls to wait with a timeout are made and
// no call to notify_* is made
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_gate), INFINITE);
assert(res == WAIT_OBJECT_0);
m_blocked -= m_gone;
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0);
assert(res);
m_gone = 0;
}
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
if (was_waiting == 1)
{
for (/**/ ; was_gone; --was_gone)
{
// better now than spurious later
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_queue), INFINITE);
assert(res == WAIT_OBJECT_0);
}
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0);
assert(res);
}
}
bool condition::do_timed_wait(const xtime& xt)
{
unsigned milliseconds;
to_duration(xt, milliseconds);
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_queue), milliseconds);
assert(res != WAIT_FAILED && res != WAIT_ABANDONED);
bool ret = (res == WAIT_OBJECT_0);
unsigned was_waiting=0;
unsigned was_gone=0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
was_waiting = m_waiting;
was_gone = m_gone;
if (was_waiting != 0)
{
if (!ret) // timeout
{
if (m_blocked != 0)
--m_blocked;
else
++m_gone; // count spurious wakeups
}
if (--m_waiting == 0)
{
if (m_blocked != 0)
{
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0); // open m_gate
assert(res);
was_waiting = 0;
}
else if (m_gone != 0)
m_gone = 0;
}
}
else if (++m_gone == (std::numeric_limits<unsigned>::max() / 2))
{
// timeout occured, normalize the m_gone count
// this may occur if many calls to wait with a timeout are made and
// no call to notify_* is made
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_gate), INFINITE);
assert(res == WAIT_OBJECT_0);
m_blocked -= m_gone;
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0);
assert(res);
m_gone = 0;
}
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
if (was_waiting == 1)
{
for (/**/ ; was_gone; --was_gone)
{
// better now than spurious later
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_queue), INFINITE);
assert(res == WAIT_OBJECT_0);
}
res = ReleaseSemaphore(reinterpret_cast<HANDLE>(m_gate), 1, 0);
assert(res);
}
return ret;
}
#elif defined(BOOST_HAS_PTHREADS)
condition::condition()
{
int res = 0;
res = pthread_cond_init(&m_condition, 0);
if (res != 0)
throw thread_resource_error();
}
condition::~condition()
{
int res = 0;
res = pthread_cond_destroy(&m_condition);
assert(res == 0);
}
void condition::notify_one()
{
int res = 0;
res = pthread_cond_signal(&m_condition);
assert(res == 0);
}
void condition::notify_all()
{
int res = 0;
res = pthread_cond_broadcast(&m_condition);
assert(res == 0);
}
void condition::do_wait(pthread_mutex_t* pmutex)
{
int res = 0;
res = pthread_cond_wait(&m_condition, pmutex);
assert(res == 0);
}
bool condition::do_timed_wait(const xtime& xt, pthread_mutex_t* pmutex)
{
timespec ts;
to_timespec(xt, ts);
int res = 0;
res = pthread_cond_timedwait(&m_condition, pmutex, &ts);
assert(res == 0 || res == ETIMEDOUT);
return res != ETIMEDOUT;
}
#endif
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 22 May 01 WEKEMPF Modified to use xtime for time outs.

405
src/mutex.cpp Normal file
View File

@@ -0,0 +1,405 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#include <boost/thread/mutex.hpp>
#include <boost/thread/xtime.hpp>
#include <boost/thread/thread.hpp>
#include <boost/thread/exceptions.hpp>
#include <boost/limits.hpp>
#include <stdexcept>
#include <cassert>
#include "timeconv.inl"
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
# include <time.h>
#elif defined(BOOST_HAS_PTHREADS)
# include <errno.h>
#endif
namespace boost {
#if defined(BOOST_HAS_WINTHREADS)
mutex::mutex()
{
m_mutex = reinterpret_cast<unsigned long>(CreateMutex(0, 0, 0));
if (!m_mutex)
throw thread_resource_error();
}
mutex::~mutex()
{
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void mutex::do_lock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
}
void mutex::do_unlock()
{
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void mutex::do_lock(cv_state&)
{
do_lock();
}
void mutex::do_unlock(cv_state&)
{
do_unlock();
}
try_mutex::try_mutex()
{
m_mutex = reinterpret_cast<unsigned long>(CreateMutex(0, 0, 0));
if (!m_mutex)
throw thread_resource_error();
}
try_mutex::~try_mutex()
{
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void try_mutex::do_lock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
}
bool try_mutex::do_trylock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), 0);
assert(res != WAIT_FAILED && res != WAIT_ABANDONED);
return res == WAIT_OBJECT_0;
}
void try_mutex::do_unlock()
{
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void try_mutex::do_lock(cv_state&)
{
do_lock();
}
void try_mutex::do_unlock(cv_state&)
{
do_unlock();
}
timed_mutex::timed_mutex()
{
m_mutex = reinterpret_cast<unsigned long>(CreateMutex(0, 0, 0));
if (!m_mutex)
throw thread_resource_error();
}
timed_mutex::~timed_mutex()
{
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void timed_mutex::do_lock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
}
bool timed_mutex::do_trylock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), 0);
assert(res != WAIT_FAILED && res != WAIT_ABANDONED);
return res == WAIT_OBJECT_0;
}
bool timed_mutex::do_timedlock(const xtime& xt)
{
unsigned milliseconds;
to_duration(xt, milliseconds);
int res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), milliseconds);
assert(res != WAIT_FAILED && res != WAIT_ABANDONED);
return res == WAIT_OBJECT_0;
}
void timed_mutex::do_unlock()
{
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void timed_mutex::do_lock(cv_state&)
{
do_lock();
}
void timed_mutex::do_unlock(cv_state&)
{
do_unlock();
}
#elif defined(BOOST_HAS_PTHREADS)
mutex::mutex()
{
int res = 0;
res = pthread_mutex_init(&m_mutex, 0);
if (res != 0)
throw thread_resource_error();
}
mutex::~mutex()
{
int res = 0;
res = pthread_mutex_destroy(&m_mutex);
assert(res == 0);
}
void mutex::do_lock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
if (res == EDEADLK) throw lock_error();
assert(res == 0);
}
void mutex::do_unlock()
{
int res = 0;
res = pthread_mutex_unlock(&m_mutex);
if (res == EPERM) throw lock_error();
assert(res == 0);
}
void mutex::do_lock(cv_state& state)
{
}
void mutex::do_unlock(cv_state& state)
{
state.pmutex = &m_mutex;
}
try_mutex::try_mutex()
{
int res = 0;
res = pthread_mutex_init(&m_mutex, 0);
if (res != 0)
throw thread_resource_error();
}
try_mutex::~try_mutex()
{
int res = 0;
res = pthread_mutex_destroy(&m_mutex);
assert(res == 0);
}
void try_mutex::do_lock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
if (res == EDEADLK) throw lock_error();
assert(res == 0);
}
bool try_mutex::do_trylock()
{
int res = 0;
res = pthread_mutex_trylock(&m_mutex);
if (res == EDEADLK) throw lock_error();
assert(res == 0 || res == EBUSY);
return res == 0;
}
void try_mutex::do_unlock()
{
int res = 0;
res = pthread_mutex_unlock(&m_mutex);
if (res == EPERM) throw lock_error();
assert(res == 0);
}
void try_mutex::do_lock(cv_state& state)
{
}
void try_mutex::do_unlock(cv_state& state)
{
state.pmutex = &m_mutex;
}
timed_mutex::timed_mutex()
: m_locked(false)
{
int res = 0;
res = pthread_mutex_init(&m_mutex, 0);
if (res != 0)
throw thread_resource_error();
res = pthread_cond_init(&m_condition, 0);
if (res != 0)
{
pthread_mutex_destroy(&m_mutex);
throw thread_resource_error();
}
}
timed_mutex::~timed_mutex()
{
assert(!m_locked);
int res = 0;
res = pthread_mutex_destroy(&m_mutex);
assert(res == 0);
res = pthread_cond_destroy(&m_condition);
assert(res == 0);
}
void timed_mutex::do_lock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
while (m_locked)
{
res = pthread_cond_wait(&m_condition, &m_mutex);
assert(res == 0);
}
assert(!m_locked);
m_locked = true;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
bool timed_mutex::do_trylock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
bool ret = false;
if (!m_locked)
{
m_locked = true;
ret = true;
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return ret;
}
bool timed_mutex::do_timedlock(const xtime& xt)
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
timespec ts;
to_timespec(xt, ts);
while (m_locked)
{
res = pthread_cond_timedwait(&m_condition, &m_mutex, &ts);
assert(res == 0 || res == ETIMEDOUT);
if (res == ETIMEDOUT)
break;
}
bool ret = false;
if (!m_locked)
{
m_locked = true;
ret = true;
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return ret;
}
void timed_mutex::do_unlock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
assert(m_locked);
m_locked = false;
res = pthread_cond_signal(&m_condition);
assert(res == 0);
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
void timed_mutex::do_lock(cv_state& state)
{
int res = 0;
while (m_locked)
{
res = pthread_cond_wait(&m_condition, &m_mutex);
assert(res == 0);
}
assert(!m_locked);
m_locked = true;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
void timed_mutex::do_unlock(cv_state& state)
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
assert(m_locked);
m_locked = false;
res = pthread_cond_signal(&m_condition);
assert(res == 0);
state.pmutex = &m_mutex;
}
#endif
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.

96
src/once.cpp Normal file
View File

@@ -0,0 +1,96 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#include <boost/thread/once.hpp>
#include <cstdio>
#include <cassert>
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
#endif
#ifdef BOOST_NO_STDC_NAMESPACE
namespace std { using ::sprintf; }
#endif
#if defined(BOOST_HAS_PTHREADS)
namespace {
pthread_key_t key;
pthread_once_t once = PTHREAD_ONCE_INIT;
typedef void (*once_callback)();
}
extern "C" {
static void key_init()
{
pthread_key_create(&key, 0);
}
static void do_once()
{
once_callback cb = reinterpret_cast<once_callback>(pthread_getspecific(key));
(*cb)();
}
}
#endif
namespace boost {
void call_once(void (*func)(), once_flag& flag)
{
#if defined(BOOST_HAS_WINTHREADS)
once_flag tmp = flag;
// Memory barrier would be needed here to prevent race conditions on some platforms with
// partial ordering.
if (!tmp)
{
char name[41];
std::sprintf(name, "2AC1A572DB6944B0A65C38C4140AF2F4%X%X", GetCurrentProcessId(), &flag);
HANDLE mutex = CreateMutex(NULL, FALSE, name);
assert(mutex != NULL);
int res = 0;
res = WaitForSingleObject(mutex, INFINITE);
assert(res == WAIT_OBJECT_0);
tmp = flag;
if (!tmp)
{
func();
tmp = true;
// Memory barrier would be needed here to prevent race conditions on some platforms
// with partial ordering.
flag = tmp;
}
res = ReleaseMutex(mutex);
assert(res);
res = CloseHandle(mutex);
assert(res);
}
#elif defined(BOOST_HAS_PTHREADS)
pthread_once(&once, &key_init);
pthread_setspecific(key, func);
pthread_once(&flag, do_once);
#endif
}
}
// Change Log:
// 1 Aug 01 WEKEMPF Initial version.

789
src/recursive_mutex.cpp Normal file
View File

@@ -0,0 +1,789 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#include <boost/thread/recursive_mutex.hpp>
#include <boost/thread/xtime.hpp>
#include <boost/thread/thread.hpp>
#include <boost/limits.hpp>
#include <stdexcept>
#include <cassert>
#include "timeconv.inl"
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
# include <time.h>
#elif defined(BOOST_HAS_PTHREADS)
# include <errno.h>
#endif
namespace boost {
#if defined(BOOST_HAS_WINTHREADS)
recursive_mutex::recursive_mutex()
: m_count(0)
{
m_mutex = reinterpret_cast<unsigned long>(CreateMutex(0, 0, 0));
if (!m_mutex)
throw thread_resource_error();
}
recursive_mutex::~recursive_mutex()
{
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void recursive_mutex::do_lock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
if (++m_count > 1)
{
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
}
void recursive_mutex::do_unlock()
{
if (--m_count == 0)
{
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
}
void recursive_mutex::do_lock(cv_state& state)
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
m_count = state;
}
void recursive_mutex::do_unlock(cv_state& state)
{
state = m_count;
m_count = 0;
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
recursive_try_mutex::recursive_try_mutex()
: m_count(0)
{
m_mutex = reinterpret_cast<unsigned long>(CreateMutex(0, 0, 0));
if (!m_mutex)
throw thread_resource_error();
}
recursive_try_mutex::~recursive_try_mutex()
{
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void recursive_try_mutex::do_lock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
if (++m_count > 1)
{
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
}
bool recursive_try_mutex::do_trylock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), 0);
assert(res != WAIT_FAILED && res != WAIT_ABANDONED);
if (res == WAIT_OBJECT_0)
{
if (++m_count > 1)
{
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
return true;
}
return false;
}
void recursive_try_mutex::do_unlock()
{
if (--m_count == 0)
{
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
}
void recursive_try_mutex::do_lock(cv_state& state)
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
m_count = state;
}
void recursive_try_mutex::do_unlock(cv_state& state)
{
state = m_count;
m_count = 0;
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
recursive_timed_mutex::recursive_timed_mutex()
: m_count(0)
{
m_mutex = reinterpret_cast<unsigned long>(CreateMutex(0, 0, 0));
if (!m_mutex)
throw thread_resource_error();
}
recursive_timed_mutex::~recursive_timed_mutex()
{
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
void recursive_timed_mutex::do_lock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
if (++m_count > 1)
{
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
}
bool recursive_timed_mutex::do_trylock()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), 0);
assert(res != WAIT_FAILED && res != WAIT_ABANDONED);
if (res == WAIT_OBJECT_0)
{
if (++m_count > 1)
{
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
return true;
}
return false;
}
bool recursive_timed_mutex::do_timedlock(const xtime& xt)
{
unsigned milliseconds;
to_duration(xt, milliseconds);
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), milliseconds);
assert(res != WAIT_FAILED && res != WAIT_ABANDONED);
if (res == WAIT_OBJECT_0)
{
if (++m_count > 1)
{
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
return true;
}
return false;
}
void recursive_timed_mutex::do_unlock()
{
if (--m_count == 0)
{
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
}
void recursive_timed_mutex::do_lock(cv_state& state)
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_mutex), INFINITE);
assert(res == WAIT_OBJECT_0);
m_count = state;
}
void recursive_timed_mutex::do_unlock(cv_state& state)
{
state = m_count;
m_count = 0;
int res = 0;
res = ReleaseMutex(reinterpret_cast<HANDLE>(m_mutex));
assert(res);
}
#elif defined(BOOST_HAS_PTHREADS)
recursive_mutex::recursive_mutex()
: m_count(0)
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
, m_valid_id(false)
# endif
{
pthread_mutexattr_t attr;
int res = 0;
res = pthread_mutexattr_init(&attr);
assert(res == 0);
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
res = pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_RECURSIVE);
assert(res == 0);
# endif
res = pthread_mutex_init(&m_mutex, &attr);
if (res != 0)
throw thread_resource_error();
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
res = pthread_cond_init(&m_unlocked, 0);
if (res != 0)
{
pthread_mutex_destroy(&m_mutex);
throw thread_resource_error();
}
# endif
}
recursive_mutex::~recursive_mutex()
{
int res = 0;
res = pthread_mutex_destroy(&m_mutex);
assert(res == 0);
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
res = pthread_cond_destroy(&m_unlocked);
assert(res == 0);
# endif
}
void recursive_mutex::do_lock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
if (++m_count > 1)
{
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
# else
pthread_t tid = pthread_self();
if (m_valid_id && pthread_equal(m_thread_id, tid))
++m_count;
else
{
while (m_valid_id)
{
res = pthread_cond_wait(&m_unlocked, &m_mutex);
assert(res == 0);
}
m_thread_id = tid;
m_valid_id = true;
m_count = 1;
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
# endif
}
void recursive_mutex::do_unlock()
{
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
if (--m_count == 0)
{
int res = 0;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
# else
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
pthread_t tid = pthread_self();
if (m_valid_id && !pthread_equal(m_thread_id, tid))
{
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
throw lock_error();
}
if (--m_count == 0)
{
assert(m_valid_id);
m_valid_id = false;
res = pthread_cond_signal(&m_unlocked);
assert(res == 0);
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
# endif
}
void recursive_mutex::do_lock(cv_state& state)
{
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
m_count = state.count;
# else
int res = 0;
while (m_valid_id)
{
res = pthread_cond_wait(&m_unlocked, &m_mutex);
assert(res == 0);
}
m_thread_id = pthread_self();
m_valid_id = true;
m_count = state.count;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
# endif
}
void recursive_mutex::do_unlock(cv_state& state)
{
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
assert(m_valid_id);
m_valid_id = false;
res = pthread_cond_signal(&m_unlocked);
assert(res == 0);
# endif
state.pmutex = &m_mutex;
state.count = m_count;
}
recursive_try_mutex::recursive_try_mutex()
: m_count(0)
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
, m_valid_id(false)
# endif
{
pthread_mutexattr_t attr;
int res = 0;
res = pthread_mutexattr_init(&attr);
assert(res == 0);
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
res = pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_RECURSIVE);
assert(res == 0);
# endif
res = pthread_mutex_init(&m_mutex, &attr);
if (res != 0)
throw thread_resource_error();
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
res = pthread_cond_init(&m_unlocked, 0);
if (res != 0)
{
pthread_mutex_destroy(&m_mutex);
throw thread_resource_error();
}
# endif
}
recursive_try_mutex::~recursive_try_mutex()
{
int res = 0;
res = pthread_mutex_destroy(&m_mutex);
assert(res == 0);
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
res = pthread_cond_destroy(&m_unlocked);
assert(res == 0);
# endif
}
void recursive_try_mutex::do_lock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
if (++m_count > 1)
{
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
# else
pthread_t tid = pthread_self();
if (m_valid_id && pthread_equal(m_thread_id, tid))
++m_count;
else
{
while (m_valid_id)
{
res = pthread_cond_wait(&m_unlocked, &m_mutex);
assert(res == 0);
}
m_thread_id = tid;
m_valid_id = true;
m_count = 1;
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
# endif
}
bool recursive_try_mutex::do_trylock()
{
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
int res = 0;
res = pthread_mutex_trylock(&m_mutex);
assert(res == 0);
if (res == 0)
{
if (++m_count > 1)
{
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
return true;
}
return false;
# else
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
bool ret = false;
pthread_t tid = pthread_self();
if (m_valid_id && pthread_equal(m_thread_id, tid))
{
++m_count;
ret = true;
}
else if (!m_valid_id)
{
m_thread_id = tid;
m_valid_id = true;
m_count = 1;
ret = true;
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return ret;
# endif
}
void recursive_try_mutex::do_unlock()
{
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
if (--m_count == 0)
{
int res = 0;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
# else
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
pthread_t tid = pthread_self();
if (m_valid_id && !pthread_equal(m_thread_id, tid))
{
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
throw lock_error();
}
if (--m_count == 0)
{
assert(m_valid_id);
m_valid_id = false;
res = pthread_cond_signal(&m_unlocked);
assert(res == 0);
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
# endif
}
void recursive_try_mutex::do_lock(cv_state& state)
{
# if defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
m_count = state.count;
# else
int res = 0;
while (m_valid_id)
{
res = pthread_cond_wait(&m_unlocked, &m_mutex);
assert(res == 0);
}
m_thread_id = pthread_self();
m_valid_id = true;
m_count = state.count;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
# endif
}
void recursive_try_mutex::do_unlock(cv_state& state)
{
# if !defined(BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE)
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
assert(m_valid_id);
m_valid_id = false;
res = pthread_cond_signal(&m_unlocked);
assert(res == 0);
# endif
state.pmutex = &m_mutex;
state.count = m_count;
}
recursive_timed_mutex::recursive_timed_mutex()
: m_valid_id(false), m_count(0)
{
int res = 0;
res = pthread_mutex_init(&m_mutex, 0);
if (res != 0)
throw thread_resource_error();
res = pthread_cond_init(&m_unlocked, 0);
if (res != 0)
{
pthread_mutex_destroy(&m_mutex);
throw thread_resource_error();
}
}
recursive_timed_mutex::~recursive_timed_mutex()
{
int res = 0;
res = pthread_mutex_destroy(&m_mutex);
assert(res == 0);
res = pthread_cond_destroy(&m_unlocked);
assert(res == 0);
}
void recursive_timed_mutex::do_lock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
pthread_t tid = pthread_self();
if (m_valid_id && pthread_equal(m_thread_id, tid))
++m_count;
else
{
while (m_valid_id)
{
res = pthread_cond_wait(&m_unlocked, &m_mutex);
assert(res == 0);
}
m_thread_id = tid;
m_valid_id = true;
m_count = 1;
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
bool recursive_timed_mutex::do_trylock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
bool ret = false;
pthread_t tid = pthread_self();
if (m_valid_id && pthread_equal(m_thread_id, tid))
{
++m_count;
ret = true;
}
else if (!m_valid_id)
{
m_thread_id = tid;
m_valid_id = true;
m_count = 1;
ret = true;
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return ret;
}
bool recursive_timed_mutex::do_timedlock(const xtime& xt)
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
bool ret = false;
pthread_t tid = pthread_self();
if (m_valid_id && pthread_equal(m_thread_id, tid))
{
++m_count;
ret = true;
}
else
{
timespec ts;
to_timespec(xt, ts);
while (m_valid_id)
{
res = pthread_cond_timedwait(&m_unlocked, &m_mutex, &ts);
if (res == ETIMEDOUT)
break;
assert(res == 0);
}
if (!m_valid_id)
{
m_thread_id = tid;
m_valid_id = true;
m_count = 1;
ret = true;
}
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return ret;
}
void recursive_timed_mutex::do_unlock()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
pthread_t tid = pthread_self();
if (m_valid_id && !pthread_equal(m_thread_id, tid))
{
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
throw lock_error();
}
if (--m_count == 0)
{
assert(m_valid_id);
m_valid_id = false;
res = pthread_cond_signal(&m_unlocked);
assert(res == 0);
}
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
void recursive_timed_mutex::do_lock(cv_state& state)
{
int res = 0;
while (m_valid_id)
{
res = pthread_cond_wait(&m_unlocked, &m_mutex);
assert(res == 0);
}
m_thread_id = pthread_self();
m_valid_id = true;
m_count = state.count;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
void recursive_timed_mutex::do_unlock(cv_state& state)
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
assert(m_valid_id);
m_valid_id = false;
res = pthread_cond_signal(&m_unlocked);
assert(res == 0);
state.pmutex = &m_mutex;
state.count = m_count;
}
#endif
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.

179
src/semaphore.cpp Normal file
View File

@@ -0,0 +1,179 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#include <boost/thread/semaphore.hpp>
#include <boost/thread/xtime.hpp>
#include <boost/limits.hpp>
#include <boost/thread/exceptions.hpp>
#include <stdexcept>
#include <cassert>
#include "timeconv.inl"
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
#elif defined(BOOST_HAS_PTHREADS)
# include <pthread.h>
# include <errno.h>
# include <boost/thread/mutex.hpp>
# include <boost/thread/condition.hpp>
#endif
namespace boost {
#if defined(BOOST_HAS_WINTHREADS)
semaphore::semaphore(unsigned count, unsigned max)
{
if (static_cast<long>(max) <= 0)
max = std::numeric_limits<long>::max();
m_sema = reinterpret_cast<unsigned long>(CreateSemaphore(0, count, max, 0));
if (!m_sema)
throw thread_resource_error();
}
semaphore::~semaphore()
{
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_sema));
assert(res);
}
bool semaphore::up(unsigned count, unsigned* prev)
{
long p;
bool ret = !!ReleaseSemaphore(reinterpret_cast<HANDLE>(m_sema), count, &p);
if (prev)
*prev = p;
return ret;
}
void semaphore::down()
{
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_sema), INFINITE);
assert(res == WAIT_OBJECT_0);
}
bool semaphore::down(const xtime& xt)
{
unsigned milliseconds;
to_duration(xt, milliseconds);
int res = 0;
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_sema), milliseconds);
assert(res != WAIT_FAILED && res != WAIT_ABANDONED);
return res == WAIT_OBJECT_0;
}
#elif defined(BOOST_HAS_PTHREADS)
semaphore::semaphore(unsigned count, unsigned max)
: m_available(count), m_max(max ? max : std::numeric_limits<unsigned>::max())
{
int res = 0;
res = pthread_mutex_init(&m_mutex, 0);
if (res != 0)
throw thread_resource_error();
res = pthread_cond_init(&m_condition, 0);
if (res != 0)
{
pthread_mutex_destroy(&m_mutex);
throw thread_resource_error();
}
}
semaphore::~semaphore()
{
int res = 0;
res = pthread_mutex_destroy(&m_mutex);
assert(res == 0);
res = pthread_cond_destroy(&m_condition);
assert(res == 0);
}
bool semaphore::up(unsigned count, unsigned* prev)
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
if (prev)
*prev = m_available;
if (m_available + count > m_max)
{
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return false;
}
m_available += count;
res = pthread_cond_broadcast(&m_condition);
assert(res == 0);
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return true;
}
void semaphore::down()
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
while (m_available == 0)
{
res = pthread_cond_wait(&m_condition, &m_mutex);
assert(res == 0);
}
m_available--;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
}
bool semaphore::down(const xtime& xt)
{
int res = 0;
res = pthread_mutex_lock(&m_mutex);
assert(res == 0);
timespec ts;
to_timespec(xt, ts);
while (m_available == 0)
{
res = pthread_cond_timedwait(&m_condition, &m_mutex, &ts);
assert(res == 0 || res == ETIMEDOUT);
if (res == ETIMEDOUT)
{
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return false;
}
}
m_available--;
res = pthread_mutex_unlock(&m_mutex);
assert(res == 0);
return true;
}
#endif
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 22 May 01 WEKEMPF Modified to use xtime for time outs.

255
src/thread.cpp Normal file
View File

@@ -0,0 +1,255 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#include <boost/thread/thread.hpp>
#include <boost/thread/semaphore.hpp>
#include <boost/thread/xtime.hpp>
#include <boost/thread/condition.hpp>
#include <cassert>
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
# include <process.h>
#endif
#include "timeconv.inl"
namespace {
class thread_param
{
public:
thread_param(const boost::function0<void>& threadfunc) : m_threadfunc(threadfunc), m_started(false) { }
void wait()
{
boost::mutex::scoped_lock scoped_lock(m_mutex);
while (!m_started)
m_condition.wait(scoped_lock);
}
void started()
{
boost::mutex::scoped_lock scoped_lock(m_mutex);
m_started = true;
m_condition.notify_one();
}
boost::mutex m_mutex;
boost::condition m_condition;
const boost::function0<void>& m_threadfunc;
bool m_started;
};
} // unnamed namespace
extern "C" {
#if defined(BOOST_HAS_WINTHREADS)
unsigned __stdcall thread_proxy(void* param)
#elif defined(BOOST_HAS_PTHREADS)
static void* thread_proxy(void* param)
#endif
{
try
{
thread_param* p = static_cast<thread_param*>(param);
boost::function0<void> threadfunc = p->m_threadfunc;
p->started();
threadfunc();
}
catch (...)
{
}
return 0;
}
}
namespace boost {
thread::thread()
: m_joinable(false)
{
#if defined(BOOST_HAS_WINTHREADS)
m_thread = reinterpret_cast<unsigned long>(GetCurrentThread());
m_id = GetCurrentThreadId();
#elif defined(BOOST_HAS_PTHREADS)
m_thread = pthread_self();
#endif
}
thread::thread(const function0<void>& threadfunc)
: m_joinable(true)
{
thread_param param(threadfunc);
#if defined(BOOST_HAS_WINTHREADS)
m_thread = _beginthreadex(0, 0, &thread_proxy, &param, 0, &m_id);
if (!m_thread)
throw thread_resource_error();
#elif defined(BOOST_HAS_PTHREADS)
int res = 0;
res = pthread_create(&m_thread, 0, &thread_proxy, &param);
if (res != 0)
throw thread_resource_error();
#endif
param.wait();
}
thread::~thread()
{
if (m_joinable)
{
#if defined(BOOST_HAS_WINTHREADS)
int res = 0;
res = CloseHandle(reinterpret_cast<HANDLE>(m_thread));
assert(res);
#elif defined(BOOST_HAS_PTHREADS)
pthread_detach(m_thread);
#endif
}
}
bool thread::operator==(const thread& other) const
{
#if defined(BOOST_HAS_WINTHREADS)
return other.m_id == m_id;
#elif defined(BOOST_HAS_PTHREADS)
return pthread_equal(m_thread, other.m_thread) != 0;
#endif
}
bool thread::operator!=(const thread& other) const
{
return !operator==(other);
}
void thread::join()
{
int res = 0;
#if defined(BOOST_HAS_WINTHREADS)
res = WaitForSingleObject(reinterpret_cast<HANDLE>(m_thread), INFINITE);
assert(res == WAIT_OBJECT_0);
res = CloseHandle(reinterpret_cast<HANDLE>(m_thread));
assert(res);
#elif defined(BOOST_HAS_PTHREADS)
res = pthread_join(m_thread, 0);
assert(res == 0);
#endif
// This isn't a race condition since any race that could occur would
// have us in undefined behavior territory any way.
m_joinable = false;
}
void thread::sleep(const xtime& xt)
{
#if defined(BOOST_HAS_WINTHREADS)
unsigned milliseconds;
to_duration(xt, milliseconds);
Sleep(milliseconds);
#elif defined(BOOST_HAS_PTHREADS)
# if defined(BOOST_HAS_PTHREAD_DELAY_NP)
timespec ts;
to_timespec(xt, ts);
int res = 0;
res = pthread_delay_np(&ts);
assert(res == 0);
# elif defined(BOOST_HAS_NANOSLEEP)
timespec ts;
to_timespec_duration(xt, ts);
// nanosleep takes a timespec that is an offset, not
// an absolute time.
nanosleep(&ts, 0);
# else
semaphore sema;
sema.down(xt);
# endif
#endif
}
void thread::yield()
{
#if defined(BOOST_HAS_WINTHREADS)
Sleep(0);
#elif defined(BOOST_HAS_PTHREADS)
# if defined(BOOST_HAS_SCHED_YIELD)
int res = 0;
res = sched_yield();
assert(res == 0);
# elif defined(BOOST_HAS_PTHREAD_YIELD)
int res = 0;
res = pthread_yield();
assert(res == 0);
# else
xtime xt;
xtime_get(&xt, TIME_UTC);
sleep(xt);
# endif
#endif
}
thread_group::thread_group()
{
}
thread_group::~thread_group()
{
// We shouldn't have to scoped_lock here, since referencing this object from another thread
// while we're deleting it in the current thread is going to lead to undefined behavior
// any way.
for (std::list<thread*>::iterator it = m_threads.begin(); it != m_threads.end(); ++it)
delete (*it);
}
thread* thread_group::create_thread(const function0<void>& threadfunc)
{
// No scoped_lock required here since the only "shared data" that's modified here occurs
// inside add_thread which does scoped_lock.
std::auto_ptr<thread> thrd(new thread(threadfunc));
add_thread(thrd.get());
return thrd.release();
}
void thread_group::add_thread(thread* thrd)
{
mutex::scoped_lock scoped_lock(m_mutex);
// For now we'll simply ignore requests to add a thread object multiple times.
// Should we consider this an error and either throw or return an error value?
std::list<thread*>::iterator it = std::find(m_threads.begin(), m_threads.end(), thrd);
assert(it == m_threads.end());
if (it == m_threads.end())
m_threads.push_back(thrd);
}
void thread_group::remove_thread(thread* thrd)
{
mutex::scoped_lock scoped_lock(m_mutex);
// For now we'll simply ignore requests to remove a thread object that's not in the group.
// Should we consider this an error and either throw or return an error value?
std::list<thread*>::iterator it = std::find(m_threads.begin(), m_threads.end(), thrd);
assert(it != m_threads.end());
if (it != m_threads.end())
m_threads.erase(it);
}
void thread_group::join_all()
{
mutex::scoped_lock scoped_lock(m_mutex);
for (std::list<thread*>::iterator it = m_threads.begin(); it != m_threads.end(); ++it)
(*it)->join();
}
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.
// 1 Jun 01 WEKEMPF Added boost::thread initial implementation.
// 3 Jul 01 WEKEMPF Redesigned boost::thread to be noncopyable.

138
src/threadmon.cpp Normal file
View File

@@ -0,0 +1,138 @@
// threadmon.cpp : Defines the entry point for the DLL application.
//
#define BOOST_THREADMON_EXPORTS
#include "threadmon.hpp"
#define WIN32_LEAN_AND_MEAN // Exclude rarely-used stuff from Windows headers
#include <windows.h>
#pragma warning(disable : 4786)
#include <list>
#include <set>
#include <algorithm>
typedef void (__cdecl * handler)(void);
typedef std::list<handler> exit_handlers;
typedef std::set<exit_handlers*> registered_handlers;
namespace
{
CRITICAL_SECTION cs;
DWORD key;
registered_handlers registry;
}
#if defined(__BORLANDC__)
#define DllMain DllEntryPoint
#endif
BOOL APIENTRY DllMain(HANDLE module, DWORD reason, LPVOID)
{
switch (reason)
{
case DLL_PROCESS_ATTACH:
InitializeCriticalSection(&cs);
key = TlsAlloc();
break;
case DLL_THREAD_ATTACH:
break;
case DLL_THREAD_DETACH:
{
// Call the thread's exit handlers.
exit_handlers* handlers = static_cast<exit_handlers*>(TlsGetValue(key));
if (handlers)
{
for (exit_handlers::iterator it = handlers->begin(); it != handlers->end(); ++it)
(*it)();
// Remove the exit handler list from the registered lists and then destroy it.
EnterCriticalSection(&cs);
registry.erase(handlers);
LeaveCriticalSection(&cs);
delete handlers;
}
}
break;
case DLL_PROCESS_DETACH:
{
// Assume the main thread is ending (call its handlers) and all other threads
// have already ended. If this DLL is loaded and unloaded dynamically at run time
// this is a bad assumption, but this is the best we can do.
exit_handlers* handlers = static_cast<exit_handlers*>(TlsGetValue(key));
if (handlers)
{
for (exit_handlers::iterator it = handlers->begin(); it != handlers->end(); ++it)
(*it)();
}
// Destroy any remaining exit handlers. Above we assumed there'd only be the main
// thread left, but to insure we don't get memory leaks we won't make that assumption
// here.
EnterCriticalSection(&cs);
for (registered_handlers::iterator it = registry.begin(); it != registry.end(); ++it)
delete (*it);
LeaveCriticalSection(&cs);
DeleteCriticalSection(&cs);
TlsFree(key);
}
break;
}
return TRUE;
}
int on_thread_exit(void (__cdecl * func)(void))
{
// Get the exit handlers for the current thread, creating and registering
// one if it doesn't exist.
exit_handlers* handlers = static_cast<exit_handlers*>(TlsGetValue(key));
if (!handlers)
{
try
{
handlers = new exit_handlers;
// Handle "broken" implementations of operator new that don't throw.
if (!handlers)
return -1;
}
catch (...)
{
return -1;
}
// Attempt to set a TLS value for the new handlers.
if (!TlsSetValue(key, handlers))
{
delete handlers;
return -1;
}
// Attempt to register this new handler so that memory can be properly
// cleaned up.
try
{
EnterCriticalSection(&cs);
registry.insert(handlers);
LeaveCriticalSection(&cs);
}
catch (...)
{
LeaveCriticalSection(&cs);
delete handlers;
return -1;
}
}
// Attempt to add the handler to the list of exit handlers. If it's been previously
// added just report success and exit.
try
{
handlers->push_front(func);
}
catch (...)
{
return -1;
}
return 0;
}

14
src/threadmon.hpp Normal file
View File

@@ -0,0 +1,14 @@
// The following ifdef block is the standard way of creating macros which make exporting
// from a DLL simpler. All files within this DLL are compiled with the BOOST_THREADMON_EXPORTS
// symbol defined on the command line. this symbol should not be defined on any project
// that uses this DLL. This way any other project whose source files include this file see
// BOOST_THREADMON_API functions as being imported from a DLL, wheras this DLL sees symbols
// defined with this macro as being exported.
#ifdef BOOST_THREADMON_EXPORTS
#define BOOST_THREADMON_API __declspec(dllexport)
#else
#define BOOST_THREADMON_API __declspec(dllimport)
#endif
extern "C" BOOST_THREADMON_API int on_thread_exit(void (__cdecl * func)(void));

92
src/timeconv.inl Normal file
View File

@@ -0,0 +1,92 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
namespace {
const unsigned MILLISECONDS_PER_SECOND = 1000;
const unsigned NANOSECONDS_PER_SECOND = 1000000000;
const unsigned NANOSECONDS_PER_MILLISECOND = 1000000;
inline void to_time(unsigned milliseconds, boost::xtime& xt)
{
int res = 0;
res = boost::xtime_get(&xt, boost::TIME_UTC);
assert(res == boost::TIME_UTC);
xt.sec += (milliseconds / MILLISECONDS_PER_SECOND);
xt.nsec += ((milliseconds % MILLISECONDS_PER_SECOND) * NANOSECONDS_PER_MILLISECOND);
if (xt.nsec > NANOSECONDS_PER_SECOND)
{
++xt.sec;
xt.nsec -= NANOSECONDS_PER_SECOND;
}
}
#if defined(BOOST_HAS_PTHREADS)
inline void to_timespec(const boost::xtime& xt, timespec& ts)
{
ts.tv_sec = static_cast<int>(xt.sec);
ts.tv_nsec = static_cast<int>(xt.nsec);
}
inline void to_time(unsigned milliseconds, timespec& ts)
{
boost::xtime xt;
to_time(milliseconds, xt);
to_timespec(xt, ts);
}
inline void to_timespec_duration(const boost::xtime& xt, timespec& ts)
{
boost::xtime cur;
int res = 0;
res = boost::xtime_get(&cur, boost::TIME_UTC);
assert(res == boost::TIME_UTC);
if (xt.sec < cur.sec || (xt.sec == cur.sec && xt.nsec < cur.nsec))
{
ts.tv_sec = 0;
ts.tv_nsec = 0;
}
else
{
ts.tv_sec = xt.sec - cur.sec;
ts.tv_nsec = xt.nsec - cur.nsec;
if( ts.tv_nsec < 0 )
{
ts.tv_sec -= 1;
ts.tv_nsec += NANOSECONDS_PER_SECOND;
}
}
}
#endif
inline void to_duration(const boost::xtime& xt, unsigned& milliseconds)
{
boost::xtime cur;
int res = 0;
res = boost::xtime_get(&cur, boost::TIME_UTC);
assert(res == boost::TIME_UTC);
if (xt.sec < cur.sec || (xt.sec == cur.sec && xt.nsec < cur.nsec))
milliseconds = 0;
else
{
milliseconds = static_cast<unsigned>(((xt.sec - cur.sec) * MILLISECONDS_PER_SECOND) +
(((xt.nsec - cur.nsec) + (NANOSECONDS_PER_MILLISECOND/2)) /
NANOSECONDS_PER_MILLISECOND));
}
}
}
// Change Log:
// 1 Jun 01 Initial creation.

145
src/tss.cpp Normal file
View File

@@ -0,0 +1,145 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#include <boost/thread/tss.hpp>
#include <boost/thread/once.hpp>
#include <boost/thread/exceptions.hpp>
#include <stdexcept>
#include <cassert>
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
#endif
#if defined(BOOST_HAS_WINTHREADS)
#include "threadmon.hpp"
#include <map>
namespace {
typedef std::pair<void(*)(void*), void*> cleanup_info;
typedef std::map<int, cleanup_info> cleanup_handlers;
DWORD key;
boost::once_flag once = boost::once_init;
void init_cleanup_key()
{
key = TlsAlloc();
assert(key != 0xFFFFFFFF);
}
void __cdecl cleanup()
{
cleanup_handlers* handlers = static_cast<cleanup_handlers*>(TlsGetValue(key));
for (cleanup_handlers::iterator it = handlers->begin(); it != handlers->end(); ++it)
{
cleanup_info info = it->second;
if (info.second)
info.first(info.second);
}
delete handlers;
}
cleanup_handlers* get_handlers()
{
boost::call_once(&init_cleanup_key, once);
cleanup_handlers* handlers = static_cast<cleanup_handlers*>(TlsGetValue(key));
if (!handlers)
{
try
{
handlers = new cleanup_handlers;
}
catch (...)
{
return 0;
}
int res = 0;
res = TlsSetValue(key, handlers);
assert(res);
res = on_thread_exit(&cleanup);
assert(res == 0);
}
return handlers;
}
}
#endif
namespace boost { namespace detail {
#if defined(BOOST_HAS_WINTHREADS)
tss::tss(void (*cleanup)(void*))
{
m_key = TlsAlloc();
if (m_key == 0xFFFFFFFF)
throw thread_resource_error();
m_cleanup = cleanup;
}
tss::~tss()
{
int res = 0;
res = TlsFree(m_key);
assert(res);
}
void* tss::get() const
{
return TlsGetValue(m_key);
}
bool tss::set(void* value)
{
if (value && m_cleanup)
{
cleanup_handlers* handlers = get_handlers();
assert(handlers);
if (!handlers)
return false;
cleanup_info info(m_cleanup, value);
(*handlers)[m_key] = info;
}
return !!TlsSetValue(m_key, value);
}
#elif defined(BOOST_HAS_PTHREADS)
tss::tss(void (*cleanup)(void*))
{
int res = 0;
res = pthread_key_create(&m_key, cleanup);
if (res != 0)
throw thread_resource_error();
}
tss::~tss()
{
int res = 0;
res = pthread_key_delete(m_key);
assert(res == 0);
}
void* tss::get() const
{
return pthread_getspecific(m_key);
}
bool tss::set(void* value)
{
return pthread_setspecific(m_key, value) == 0;
}
#endif
} // namespace detail
} // namespace boost
// Change Log:
// 6 Jun 01 WEKEMPF Initial version.

56
src/xtime.cpp Normal file
View File

@@ -0,0 +1,56 @@
// Copyright (C) 2001
// William E. Kempf
//
// Permission to use, copy, modify, distribute and sell this software
// and its documentation for any purpose is hereby granted without fee,
// provided that the above copyright notice appear in all copies and
// that both that copyright notice and this permission notice appear
// in supporting documentation. William E. Kempf makes no representations
// about the suitability of this software for any purpose.
// It is provided "as is" without express or implied warranty.
#include <boost/thread/xtime.hpp>
#if defined(BOOST_HAS_FTIME)
# include <windows.h>
#elif defined(BOOST_HAS_GETTIMEOFDAY)
# include <sys/time.h>
#endif
namespace boost {
int xtime_get(struct xtime* xtp, int clock_type)
{
if (clock_type == TIME_UTC)
{
#if defined(BOOST_HAS_FTIME)
FILETIME ft;
GetSystemTimeAsFileTime(&ft);
const __int64 TIMESPEC_TO_FILETIME_OFFSET = ((__int64)27111902 << 32) + (__int64)3577643008;
xtp->sec = (int)((*(__int64*)&ft - TIMESPEC_TO_FILETIME_OFFSET) / 10000000);
xtp->nsec = (int)((*(__int64*)&ft - TIMESPEC_TO_FILETIME_OFFSET -
((__int64)xtp->sec * (__int64)10000000)) * 100);
return clock_type;
#elif defined(BOOST_HAS_GETTIMEOFDAY)
struct timeval tv;
gettimeofday(&tv, 0);
xtp->sec = tv.tv_sec;
xtp->nsec = tv.tv_usec * 1000;
return clock_type;
#elif defined(BOOST_HAS_CLOCK_GETTIME)
timespec ts;
clock_gettime(CLOCK_REALTIME, &ts);
xtp->sec = ts.tv_sec;
xtp->nsec = ts.tv_nsec;
return clock_type;
#else
# error "xtime_get implementation undefined"
#endif
}
return 0;
}
} // namespace boost
// Change Log:
// 8 Feb 01 WEKEMPF Initial version.

View File

487
test/test_thread.cpp Normal file
View File

@@ -0,0 +1,487 @@
#include <list>
#include <boost/thread/mutex.hpp>
#include <boost/thread/recursive_mutex.hpp>
#include <boost/thread/condition.hpp>
#include <boost/thread/semaphore.hpp>
//#include <boost/thread/atomic.hpp>
#include <boost/thread/tss.hpp>
#include <boost/thread/once.hpp>
#include <boost/thread/thread.hpp>
#include <boost/thread/xtime.hpp>
#define BOOST_INCLUDE_MAIN
#include <boost/test/test_tools.hpp>
#if defined(BOOST_HAS_WINTHREADS)
# include <windows.h>
#endif
template <typename M>
void test_lock(M* dummy=0)
{
typedef M mutex_type;
typedef typename M::scoped_lock lock_type;
mutex_type mutex;
boost::condition condition;
// Test the lock's constructors.
{
lock_type lock(mutex, false);
BOOST_TEST(!lock);
}
lock_type lock(mutex);
BOOST_TEST(lock);
// Construct and initialize an xtime for a fast time out.
boost::xtime xt;
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.nsec += 100000000;
// Test the lock and the mutex with condition variables.
// No one is going to notify this condition variable. We expect to
// time out.
BOOST_TEST(condition.timed_wait(lock, xt) == false);
BOOST_TEST(lock);
// Test the lock and unlock methods.
lock.unlock();
BOOST_TEST(!lock);
lock.lock();
BOOST_TEST(lock);
}
template <typename M>
void test_trylock(M* dummy=0)
{
typedef M mutex_type;
typedef typename M::scoped_try_lock try_lock_type;
mutex_type mutex;
boost::condition condition;
// Test the lock's constructors.
{
try_lock_type lock(mutex);
BOOST_TEST(lock);
}
{
try_lock_type lock(mutex, false);
BOOST_TEST(!lock);
}
try_lock_type lock(mutex, true);
BOOST_TEST(lock);
// Construct and initialize an xtime for a fast time out.
boost::xtime xt;
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.nsec += 100000000;
// Test the lock and the mutex with condition variables.
// No one is going to notify this condition variable. We expect to
// time out.
BOOST_TEST(condition.timed_wait(lock, xt) == false);
BOOST_TEST(lock);
// Test the lock, unlock and trylock methods.
lock.unlock();
BOOST_TEST(!lock);
lock.lock();
BOOST_TEST(lock);
lock.unlock();
BOOST_TEST(!lock);
BOOST_TEST(lock.try_lock());
BOOST_TEST(lock);
}
template <typename M>
void test_timedlock(M* dummy=0)
{
typedef M mutex_type;
typedef typename M::scoped_timed_lock timed_lock_type;
mutex_type mutex;
boost::condition condition;
// Test the lock's constructors.
{
// Construct and initialize an xtime for a fast time out.
boost::xtime xt;
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.nsec += 100000000;
timed_lock_type lock(mutex, xt);
BOOST_TEST(lock);
}
{
timed_lock_type lock(mutex, false);
BOOST_TEST(!lock);
}
timed_lock_type lock(mutex, true);
BOOST_TEST(lock);
// Construct and initialize an xtime for a fast time out.
boost::xtime xt;
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.nsec += 100000000;
// Test the lock and the mutex with condition variables.
// No one is going to notify this condition variable. We expect to
// time out.
BOOST_TEST(condition.timed_wait(lock, xt) == false);
BOOST_TEST(lock);
// Test the lock, unlock and timedlock methods.
lock.unlock();
BOOST_TEST(!lock);
lock.lock();
BOOST_TEST(lock);
lock.unlock();
BOOST_TEST(!lock);
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.nsec += 100000000;
BOOST_TEST(lock.timed_lock(xt));
}
void test_mutex()
{
typedef boost::mutex mutex;
test_lock<mutex>();
}
void test_try_mutex()
{
typedef boost::try_mutex mutex;
test_lock<mutex>();
test_trylock<mutex>();
}
void test_timed_mutex()
{
typedef boost::timed_mutex mutex;
test_lock<mutex>();
test_trylock<mutex>();
test_timedlock<mutex>();
}
void test_recursive_mutex()
{
typedef boost::recursive_mutex mutex;
test_lock<mutex>();
mutex mx;
mutex::scoped_lock lock1(mx);
mutex::scoped_lock lock2(mx);
}
void test_recursive_try_mutex()
{
typedef boost::recursive_try_mutex mutex;
test_lock<mutex>();
test_trylock<mutex>();
mutex mx;
mutex::scoped_lock lock1(mx);
mutex::scoped_lock lock2(mx);
}
void test_recursive_timed_mutex()
{
typedef boost::recursive_timed_mutex mutex;
test_lock<mutex>();
test_trylock<mutex>();
test_timedlock<mutex>();
mutex mx;
mutex::scoped_lock lock1(mx);
mutex::scoped_lock lock2(mx);
}
struct condition_test_data
{
condition_test_data() : notified(0), awoken(0) { }
boost::mutex mutex;
boost::condition condition;
int notified;
int awoken;
};
void condition_test_thread(void* param)
{
condition_test_data* data = static_cast<condition_test_data*>(param);
boost::mutex::scoped_lock lock(data->mutex);
BOOST_TEST(lock);
while (!(data->notified > 0))
data->condition.wait(lock);
BOOST_TEST(lock);
data->awoken++;
}
class thread_adapter
{
public:
thread_adapter(void (*func)(void*), void* param) : _func(func), _param(param) { }
void operator()() const { _func(_param); }
private:
void (*_func)(void*);
void* _param;
};
void test_condition_notify_one()
{
condition_test_data data;
boost::thread thread(thread_adapter(&condition_test_thread, &data));
{
boost::mutex::scoped_lock lock(data.mutex);
BOOST_TEST(lock);
data.notified++;
data.condition.notify_one();
}
thread.join();
BOOST_TEST(data.awoken == 1);
}
void test_condition_notify_all()
{
const int NUMTHREADS = 5;
boost::thread_group threads;
condition_test_data data;
for (int i = 0; i < NUMTHREADS; ++i)
threads.create_thread(thread_adapter(&condition_test_thread, &data));
{
boost::mutex::scoped_lock lock(data.mutex);
BOOST_TEST(lock);
data.notified++;
data.condition.notify_all();
}
threads.join_all();
BOOST_TEST(data.awoken == NUMTHREADS);
}
struct cond_predicate
{
cond_predicate(int& var, int val) : _var(var), _val(val) { }
bool operator()() { return _var == _val; }
int& _var;
int _val;
};
void condition_test_waits(void* param)
{
condition_test_data* data = static_cast<condition_test_data*>(param);
boost::mutex::scoped_lock lock(data->mutex);
BOOST_TEST(lock);
// Test wait.
while (data->notified != 1)
data->condition.wait(lock);
BOOST_TEST(lock);
BOOST_TEST(data->notified == 1);
data->awoken++;
data->condition.notify_one();
// Test predicate wait.
data->condition.wait(lock, cond_predicate(data->notified, 2));
BOOST_TEST(lock);
BOOST_TEST(data->notified == 2);
data->awoken++;
data->condition.notify_one();
// Test timed_wait.
boost::xtime xt;
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.nsec += 100000000;
while (data->notified != 3)
data->condition.timed_wait(lock, xt);
BOOST_TEST(lock);
BOOST_TEST(data->notified == 3);
data->awoken++;
data->condition.notify_one();
// Test predicate timed_wait.
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.sec += 2;
BOOST_TEST(data->condition.timed_wait(lock, xt, cond_predicate(data->notified, 4)));
BOOST_TEST(lock);
BOOST_TEST(data->notified == 4);
data->awoken++;
}
void test_condition_waits()
{
condition_test_data data;
boost::thread thread(thread_adapter(&condition_test_waits, &data));
boost::xtime xt;
{
boost::mutex::scoped_lock lock(data.mutex);
BOOST_TEST(lock);
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.sec += 1;
boost::thread::sleep(xt);
data.notified++;
data.condition.notify_one();
while (data.awoken != 1)
data.condition.wait(lock);
BOOST_TEST(data.awoken == 1);
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.sec += 1;
boost::thread::sleep(xt);
data.notified++;
data.condition.notify_one();
while (data.awoken != 2)
data.condition.wait(lock);
BOOST_TEST(data.awoken == 2);
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.sec += 1;
boost::thread::sleep(xt);
data.notified++;
data.condition.notify_one();
while (data.awoken != 3)
data.condition.wait(lock);
BOOST_TEST(data.awoken == 3);
}
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.sec += 1;
boost::thread::sleep(xt);
data.notified++;
data.condition.notify_one();
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.sec += 1;
boost::thread::sleep(xt);
thread.join();
BOOST_TEST(data.awoken == 4);
}
void test_condition()
{
test_condition_notify_one();
test_condition_notify_all();
test_condition_waits();
}
void test_semaphore()
{
boost::xtime xt;
unsigned val;
boost::semaphore sema(0, 1);
BOOST_TEST(sema.up(1, &val));
BOOST_TEST(val == 0);
BOOST_TEST(!sema.up());
sema.down();
BOOST_TEST(sema.up());
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.nsec += 100000000;
BOOST_TEST(sema.down(xt));
BOOST_TEST(boost::xtime_get(&xt, boost::TIME_UTC) == boost::TIME_UTC);
xt.nsec += 100000000;
BOOST_TEST(!sema.down(xt));
}
/*void test_atomic_t()
{
boost::atomic_t atomic;
BOOST_TEST(boost::increment(atomic) > 0);
BOOST_TEST(boost::decrement(atomic) == 0);
BOOST_TEST(boost::swap(atomic, 10) == 0);
BOOST_TEST(boost::swap(atomic, 0) == 10);
BOOST_TEST(boost::compare_swap(atomic, 20, 10) == 0);
BOOST_TEST(boost::compare_swap(atomic, 20, 0) == 0);
BOOST_TEST(boost::read(atomic) == 20);
}*/
boost::mutex tss_mutex;
int tss_instances = 0;
struct tss_value_t
{
tss_value_t()
{
boost::mutex::scoped_lock lock(tss_mutex);
++tss_instances;
value = 0;
}
~tss_value_t()
{
boost::mutex::scoped_lock lock(tss_mutex);
--tss_instances;
}
int value;
};
boost::thread_specific_ptr<tss_value_t> tss_value;
void test_tss_thread()
{
tss_value.reset(new tss_value_t());
for (int i=0; i<1000; ++i)
{
int& n = tss_value->value;
BOOST_TEST(n == i);
++n;
}
}
void test_tss()
{
const int NUMTHREADS=5;
boost::thread_group threads;
for (int i=0; i<NUMTHREADS; ++i)
threads.create_thread(&test_tss_thread);
threads.join_all();
BOOST_TEST(tss_instances == 0);
}
int once_value = 0;
boost::once_flag once = boost::once_init;
void init_once_value()
{
once_value++;
}
void test_once_thread()
{
boost::call_once(&init_once_value, once);
}
void test_once()
{
const int NUMTHREADS=5;
boost::thread_group threads;
for (int i=0; i<NUMTHREADS; ++i)
threads.create_thread(&test_once_thread);
threads.join_all();
BOOST_TEST(once_value == 1);
}
int test_main(int, char*[])
{
test_mutex();
test_try_mutex();
test_timed_mutex();
test_recursive_mutex();
test_recursive_try_mutex();
test_recursive_timed_mutex();
test_condition();
test_semaphore();
test_tss();
test_once();
return 0;
}