diff options
Diffstat (limited to 'libstdc++-v3/doc/html/manual/bk01pt03ch28s02.html')
| -rw-r--r-- | libstdc++-v3/doc/html/manual/bk01pt03ch28s02.html | 41 |
1 files changed, 41 insertions, 0 deletions
diff --git a/libstdc++-v3/doc/html/manual/bk01pt03ch28s02.html b/libstdc++-v3/doc/html/manual/bk01pt03ch28s02.html new file mode 100644 index 000000000..4350e20b1 --- /dev/null +++ b/libstdc++-v3/doc/html/manual/bk01pt03ch28s02.html @@ -0,0 +1,41 @@ +<?xml version="1.0" encoding="UTF-8" standalone="no"?> +<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd"> +<html xmlns="http://www.w3.org/1999/xhtml"><head><title>Implementation</title><meta name="generator" content="DocBook XSL-NS Stylesheets V1.76.1"/><meta name="keywords" content=" ISO C++ , library "/><meta name="keywords" content=" ISO C++ , library "/><link rel="home" href="../spine.html" title="The GNU C++ Library"/><link rel="up" href="ext_concurrency.html" title="Chapter 28. Concurrency"/><link rel="prev" href="ext_concurrency.html" title="Chapter 28. Concurrency"/><link rel="next" href="bk01pt03ch28s03.html" title="Use"/></head><body><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="3" align="center">Implementation</th></tr><tr><td align="left"><a accesskey="p" href="ext_concurrency.html">Prev</a> </td><th width="60%" align="center">Chapter 28. Concurrency</th><td align="right"> <a accesskey="n" href="bk01pt03ch28s03.html">Next</a></td></tr></table><hr/></div><div class="section" title="Implementation"><div class="titlepage"><div><div><h2 class="title"><a id="manual.ext.concurrency.impl"/>Implementation</h2></div></div></div><div class="section" title="Using Builtin Atomic Functions"><div class="titlepage"><div><div><h3 class="title"><a id="manual.ext.concurrency.impl.atomic_fallbacks"/>Using Builtin Atomic Functions</h3></div></div></div><p>The functions for atomic operations described above are either +implemented via compiler intrinsics (if the underlying host is +capable) or by library fallbacks.</p><p>Compiler intrinsics (builtins) are always preferred. However, as +the compiler builtins for atomics are not universally implemented, +using them directly is problematic, and can result in undefined +function calls. (An example of an undefined symbol from the use +of <code class="code">__sync_fetch_and_add</code> on an unsupported host is a +missing reference to <code class="code">__sync_fetch_and_add_4</code>.) +</p><p>In addition, on some hosts the compiler intrinsics are enabled +conditionally, via the <code class="code">-march</code> command line flag. This makes +usage vary depending on the target hardware and the flags used during +compile. +</p><p> +If builtins are possible for bool-sized integral types, +<code class="code">_GLIBCXX_ATOMIC_BUILTINS_1</code> will be defined. +If builtins are possible for int-sized integral types, +<code class="code">_GLIBCXX_ATOMIC_BUILTINS_4</code> will be defined. +</p><p>For the following hosts, intrinsics are enabled by default. +</p><div class="itemizedlist"><ul class="itemizedlist"><li class="listitem"><p>alpha</p></li><li class="listitem"><p>ia64</p></li><li class="listitem"><p>powerpc</p></li><li class="listitem"><p>s390</p></li></ul></div><p>For others, some form of <code class="code">-march</code> may work. On +non-ancient x86 hardware, <code class="code">-march=native</code> usually does the +trick.</p><p> For hosts without compiler intrinsics, but with capable +hardware, hand-crafted assembly is selected. This is the case for the following hosts: +</p><div class="itemizedlist"><ul class="itemizedlist"><li class="listitem"><p>cris</p></li><li class="listitem"><p>hppa</p></li><li class="listitem"><p>i386</p></li><li class="listitem"><p>i486</p></li><li class="listitem"><p>m48k</p></li><li class="listitem"><p>mips</p></li><li class="listitem"><p>sparc</p></li></ul></div><p>And for the rest, a simulated atomic lock via pthreads. +</p><p> Detailed information about compiler intrinsics for atomic operations can be found in the GCC <a class="link" href="http://gcc.gnu.org/onlinedocs/gcc/Atomic-Builtins.html"> documentation</a>. +</p><p> More details on the library fallbacks from the porting <a class="link" href="internals.html#internals.thread_safety" title="Thread Safety">section</a>. +</p></div><div class="section" title="Thread Abstraction"><div class="titlepage"><div><div><h3 class="title"><a id="manual.ext.concurrency.impl.thread"/>Thread Abstraction</h3></div></div></div><p>A thin layer above IEEE 1003.1 (i.e. pthreads) is used to abstract +the thread interface for GCC. This layer is called "gthread," and is +comprised of one header file that wraps the host's default thread layer with +a POSIX-like interface. +</p><p> The file <gthr-default.h> points to the deduced wrapper for +the current host. In libstdc++ implementation files, +<bits/gthr.h> is used to select the proper gthreads file. +</p><p>Within libstdc++ sources, all calls to underlying thread functionality +use this layer. More detail as to the specific interface can be found in the source <a class="link" href="http://gcc.gnu.org/onlinedocs/libstdc++/latest-doxygen/a00883_source.html">documentation</a>. +</p><p>By design, the gthread layer is interoperable with the types, +functions, and usage found in the usual <pthread.h> file, +including <code class="code">pthread_t</code>, <code class="code">pthread_once_t</code>, <code class="code">pthread_create</code>, +etc. +</p></div></div><div class="navfooter"><hr/><table width="100%" summary="Navigation footer"><tr><td align="left"><a accesskey="p" href="ext_concurrency.html">Prev</a> </td><td align="center"><a accesskey="u" href="ext_concurrency.html">Up</a></td><td align="right"> <a accesskey="n" href="bk01pt03ch28s03.html">Next</a></td></tr><tr><td align="left" valign="top">Chapter 28. Concurrency </td><td align="center"><a accesskey="h" href="../spine.html">Home</a></td><td align="right" valign="top"> Use</td></tr></table></div></body></html> |