boost/thread
2
0
Fork 0
mirror of https://github.com/boostorg/thread.git synced 2026-02-09 11:32:12 +00:00
Code Issues Projects Releases Wiki Activity
Files
d27e195ca243aec436afecfe6b18e5122ae53334
thread/doc/thread_formal_definition.html
Beman Dawes 5cb2a46ae4 Minor wording improvements, more references added 
[SVN r10555]
2001-07-06 19:23:45 +00:00

247 lines
9.8 KiB
HTML
Raw Blame History

<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, Formal Definition of &quot;Thread&quot;</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">Formal Definition of &quot;Thread&quot;</h2>
</td>
</tr>
</table>
<h2>Introduction</h2>
<p>The definition is given in terms of the <a href="#C++"> C++ Standard</a>.&nbsp; References to the standard
are in the form [1.2.3/4], which
represents the section number and following the &quot;/&quot;, the paragraph
number.&nbsp; A short <a href="#Bibliography">bibliography</a> is also supplied.</p>
<p>Because the definition is written in something akin to
&quot;standardese&quot;, it 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>Definition of &quot;Thread&quot;</h2>
<p>A thread 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 at least one initial thread even if the program makes no thread
library calls.</p>
<p>Each thread shares certain aspects of its execution environment with all
other threads in the program:</p>
<ul>
<li>Static storage duration (static, extern) objects [3.7.1].</li>
</ul>
<ul>
<li>Dynamic storage duration (heap) objects [3.7.3].</li>
</ul>
<ul>
<li>Unless otherwise specified, 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>Unique thread identifier.</li>
</ul>
<ul>
<li>Priority and other specified thread attributes.&nbsp;</li>
</ul>
<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>
<p>During the lifetime of a thread, it shall always 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 of the number of processors.&nbsp;</td>
</tr>
<tr>
<td>Blocked</td>
<td>Waiting for some resource other than a processor that is not currently
available, or for the completion of calls to library functions [1.9/6].</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>
<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] can be accessed from two or more threads if it is of
static storage duration (static, extern) [3.7.1], or if a pointer or reference to the
object is explicitly or
implicitly dereferenced.</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 as specified in the following
table.&nbsp;&nbsp; For the same row of the table, the value of an object as if
accessed by thread A at the indicated sequence point will be determinate and the
same if accessed at or after the indicated thread B sequence point, unless
modified by any thread between the time the thread A sequence point is reached
and thread B sequence point of access. 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>
<p>(That isn't right yet.&nbsp; What if some other thread modifies the object
after the first thread B access, then thread B accesses it again?)</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 the same as when there is no matching exception handler
in the 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>
<h2><a name="Bibliography">Bibliography</a></h2>
<p>ISO/IEC 14882:1998(E) Programming Language <a name="C++"> C++</a></p>
<blockquote>
<p>This is the official C++ Standard
document. Available from <a href="http://www.ansi.org">ANSI</a> (American
National Standards Institute) Electronic Standards Store.&nbsp;</p>
</blockquote>
<p><a href="http://cseng.aw.com/book/0,3828,0201633922,00.html">Programming with
POSIX Threads</a>, David R. Butenhof, Addison-Wesley 1997, ISBN 0-201-63392-2</p>
<blockquote>
<p>This is a superior explanation of threads and how to use them.&nbsp; Many
of the insights apply to all multi-threaded programming, not just POSIX
Threads.</p>
</blockquote>
<p><a href="http://www.acm.org/classics/feb96/&nbsp;">Monitors: An Operating System Structuring Concept</a>,
C.A.R Hoare, <cite>Communications of the ACM</cite>, Vol. 17, No. 10. October
1974, pp. 549-557</p>
<blockquote>
<p>Hoare and Brinch Hansen's work on Monitors is the basis for reliable
multi-threading patterns.</p>
</blockquote>
<p><a href="http://www.cs.wustl.edu/~schmidt/win32-cv-1.html">Strategies for
Implementing POSIX Condition Variables on Win32</a>, Douglas C. Schmidt and Irfan Pyarali,
Department of Computer Science, Washington University, St. Louis, Missouri.</p>
<blockquote>
<p>Rationale for understanding Boost.Threads condition variables. Note that Alexander Terekhov found some bugs in
the implementation used in this article, so what pthreads-win32 and Boost.Threads use is more complicated yet.</p>
</blockquote>
<hr>
<p>© Copyright Beman Dawes, 2001</p>
<p>Revised <!--webbot bot="Timestamp" S-Type="EDITED" S-Format="%d %b %Y" startspan -->06 Jul 2001<!--webbot bot="Timestamp" endspan i-checksum="14992" -->
</p>
<p>&nbsp;</p>
<p>&nbsp;</p>
<p>&nbsp;</p>
</body>
</html>
Reference in New Issue View Git Blame Copy Permalink