<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>Debugging Support</title><meta name="generator" content="DocBook XSL Stylesheets Vsnapshot" /><meta name="keywords" content="C++, debug" /><meta name="keywords" content="ISO C++, library" /><meta name="keywords" content="ISO C++, runtime, library" /><link rel="home" href="../index.html" title="The GNU C++ Library" /><link rel="up" href="using.html" title="Chapter 3. Using" /><link rel="prev" href="using_exceptions.html" title="Exceptions" /><link rel="next" href="std_contents.html" title="Part II. Standard Contents" /></head><body><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="3" align="center">Debugging Support</th></tr><tr><td width="20%" align="left"><a accesskey="p" href="using_exceptions.html">Prev</a> </td><th width="60%" align="center">Chapter 3. Using</th><td width="20%" align="right"> <a accesskey="n" href="std_contents.html">Next</a></td></tr></table><hr /></div><div class="section"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="manual.intro.using.debug"></a>Debugging Support</h2></div></div></div><p>
There are numerous things that can be done to improve the ease with
which C++ binaries are debugged when using the GNU tool chain. Here
are some of them.
</p><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="debug.compiler"></a>Using <span class="command"><strong>g++</strong></span></h3></div></div></div><p>
Compiler flags determine how debug information is transmitted
between compilation and debug or analysis tools.
</p><p>
The default optimizations and debug flags for a libstdc++ build
are <code class="code">-g -O2</code>. However, both debug and optimization
flags can be varied to change debugging characteristics. For
instance, turning off all optimization via the <code class="code">-g -O0
-fno-inline</code> flags will disable inlining and optimizations,
and include debugging information, so that stepping through all functions,
(including inlined constructors and destructors) is possible. In
addition, <code class="code">-fno-eliminate-unused-debug-types</code> can be
used when additional debug information, such as nested class info,
is desired.
</p><p>
Or, the debug format that the compiler and debugger use to
communicate information about source constructs can be changed via
<code class="code">-gdwarf-2</code> or <code class="code">-gstabs</code> flags: some debugging
formats permit more expressive type and scope information to be
shown in GDB. Expressiveness can be enhanced by flags like
<code class="code">-g3</code>. The default debug information for a particular
platform can be identified via the value set by the
PREFERRED_DEBUGGING_TYPE macro in the GCC sources.
</p><p>
Many other options are available: please see <a class="link" href="https://gcc.gnu.org/onlinedocs/gcc/Debugging-Options.html#Debugging%20Options" target="_top">"Options
for Debugging Your Program"</a> in Using the GNU Compiler
Collection (GCC) for a complete list.
</p></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="debug.debug_mode"></a>Debug Mode</h3></div></div></div><p>
The <a class="link" href="debug_mode.html" title="Chapter 17. Debug Mode">Debug Mode</a>
has compile and run-time checks for many containers.
</p><p>
There are also lightweight assertions for checking function preconditions,
such as checking for out-of-bounds indices when accessing a
<code class="classname">std::vector</code>. These can be enabled without using
the full Debug Mode, by using <code class="option">-D_GLIBCXX_ASSERTIONS</code>
(see <a class="xref" href="using_macros.html" title="Macros">Macros</a>).
</p></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="debug.exceptions"></a>Tracking uncaught exceptions</h3></div></div></div><p>
The <a class="link" href="termination.html#support.termination.verbose" title="Verbose Terminate Handler">verbose
termination handler</a> gives information about uncaught
exceptions which kill the program.
</p></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="debug.memory"></a>Memory Leak Hunting</h3></div></div></div><p>
On many targets GCC supports AddressSanitizer, a fast memory error detector,
which is enabled by the <code class="option">-fsanitize=address</code> option.
</p><p>
The <code class="classname">std::vector</code> implementation has additional
instrumentation to work with AddressSanitizer, but this has to be enabled
explicitly by using <code class="option">-D_GLIBCXX_SANITIZE_VECTOR</code>
(see <a class="xref" href="using_macros.html" title="Macros">Macros</a>).
</p><p>
There are also various third party memory tracing and debug utilities
that can be used to provide detailed memory allocation information
about C++ code. An exhaustive list of tools is not going to be
attempted, but includes <code class="code">mtrace</code>, <code class="code">valgrind</code>,
<code class="code">mudflap</code> (no longer supported since GCC 4.9.0), ElectricFence,
and the non-free commercial product <code class="code">purify</code>.
In addition, <code class="code">libcwd</code>, jemalloc and TCMalloc have replacements
for the global <code class="code">new</code> and <code class="code">delete</code> operators
that can track memory allocation and deallocation and provide useful
memory statistics.
</p><p>
For valgrind, there are some specific items to keep in mind. First
of all, use a version of valgrind that will work with current GNU
C++ tools: the first that can do this is valgrind 1.0.4, but later
versions should work better. Second, using an unoptimized build
might avoid confusing valgrind.
</p><p>
Third, it may be necessary to force deallocation in other libraries
as well, namely the "C" library. On GNU/Linux, this can be accomplished
with the appropriate use of the <code class="code">__cxa_atexit</code> or
<code class="code">atexit</code> functions.
</p><pre class="programlisting">
#include <cstdlib>
extern "C" void __libc_freeres(void);
void do_something() { }
int main()
{
atexit(__libc_freeres);
do_something();
return 0;
}
</pre><p>or, using <code class="code">__cxa_atexit</code>:</p><pre class="programlisting">
extern "C" void __libc_freeres(void);
extern "C" int __cxa_atexit(void (*func) (void *), void *arg, void *d);
void do_something() { }
int main()
{
extern void* __dso_handle __attribute__ ((__weak__));
__cxa_atexit((void (*) (void *)) __libc_freeres, NULL,
&__dso_handle ? __dso_handle : NULL);
do_test();
return 0;
}
</pre><p>
Suggested valgrind flags, given the suggestions above about setting
up the runtime environment, library, and test file, might be:
</p><pre class="programlisting">
valgrind -v --num-callers=20 --leak-check=yes --leak-resolution=high --show-reachable=yes a.out
</pre><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="debug.memory.mtalloc"></a>Non-memory leaks in Pool and MT allocators</h4></div></div></div><p>
There are different kinds of allocation schemes that can be used by
<code class="code">std::allocator</code>. Prior to GCC 3.4.0 the default was to use
a pooling allocator, <code class="classname">pool_allocator</code>,
which is still available as the optional
<code class="classname">__pool_alloc</code> extension.
Another optional extension, <code class="classname">__mt_alloc</code>,
is a high-performance pool allocator.
</p><p>
In a suspect executable these pooling allocators can give
the mistaken impression that memory is being leaked,
when in reality the memory "leak" is a pool being used
by the library's allocator and is reclaimed after program
termination.
</p><p>
If you're using memory debugging tools on a program that uses
one of these pooling allocators, you can set the environment variable
<code class="literal">GLIBCXX_FORCE_NEW</code> to keep extraneous pool allocation
noise from cluttering debug information.
For more details, see the
<a class="link" href="mt_allocator.html" title="Chapter 19. The mt_allocator">mt allocator</a>
documentation and look specifically for <code class="code">GLIBCXX_FORCE_NEW</code>.
</p></div></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="debug.races"></a>Data Race Hunting</h3></div></div></div><p>
All synchronization primitives used in the library internals need to be
understood by race detectors so that they do not produce false reports.
</p><p>
Two annotation macros are used to explain low-level synchronization
to race detectors:
<code class="code">_GLIBCXX_SYNCHRONIZATION_HAPPENS_BEFORE()</code> and
<code class="code"> _GLIBCXX_SYNCHRONIZATION_HAPPENS_AFTER()</code>.
By default, these macros are defined empty -- anyone who wants
to use a race detector needs to redefine them to call an
appropriate API.
Since these macros are empty by default when the library is built,
redefining them will only affect inline functions and template
instantiations which are compiled in user code. This allows annotation
of templates such as <code class="code">shared_ptr</code>, but not code which is
only instantiated in the library. Code which is only instantiated in
the library needs to be recompiled with the annotation macros defined.
That can be done by rebuilding the entire
<code class="filename">libstdc++.so</code> file but a simpler
alternative exists for ELF platforms such as GNU/Linux, because ELF
symbol interposition allows symbols defined in the shared library to be
overridden by symbols with the same name that appear earlier in the
runtime search path. This means you only need to recompile the functions
that are affected by the annotation macros, which can be done by
recompiling individual files.
Annotating <code class="code">std::string</code> and <code class="code">std::wstring</code>
reference counting can be done by disabling extern templates (by defining
<code class="code">_GLIBCXX_EXTERN_TEMPLATE=-1</code>) or by rebuilding the
<code class="filename">src/string-inst.cc</code> file.
Annotating the remaining atomic operations (at the time of writing these
are in <code class="code">ios_base::Init::~Init</code>, <code class="code">locale::_Impl</code>,
<code class="code">locale::facet</code> and <code class="code">thread::_M_start_thread</code>)
requires rebuilding the relevant source files.
</p><p>
The approach described above is known to work with the following race
detection tools:
<a class="link" href="https://valgrind.org/docs/manual/drd-manual.html" target="_top">
DRD</a>,
<a class="link" href="https://valgrind.org/docs/manual/hg-manual.html" target="_top">
Helgrind</a>, and
<a class="link" href="https://github.com/google/sanitizers" target="_top">
ThreadSanitizer</a> (this refers to ThreadSanitizer v1, not the
new "tsan" feature built-in to GCC itself).
</p><p>
With DRD, Helgrind and ThreadSanitizer you will need to define
the macros like this:
</p><pre class="programlisting">
#define _GLIBCXX_SYNCHRONIZATION_HAPPENS_BEFORE(A) ANNOTATE_HAPPENS_BEFORE(A)
#define _GLIBCXX_SYNCHRONIZATION_HAPPENS_AFTER(A) ANNOTATE_HAPPENS_AFTER(A)
</pre><p>
Refer to the documentation of each particular tool for details.
</p></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="debug.gdb"></a>Using <span class="command"><strong>gdb</strong></span></h3></div></div></div><p>
</p><p>
Many options are available for GDB itself: please see <a class="link" href="https://sourceware.org/gdb/current/onlinedocs/gdb" target="_top">
"GDB features for C++" </a> in the GDB documentation. Also
recommended: the other parts of this manual.
</p><p>
These settings can either be switched on in at the GDB command line,
or put into a <code class="filename">.gdbinit</code> file to establish default
debugging characteristics, like so:
</p><pre class="programlisting">
set print pretty on
set print object on
set print static-members on
set print vtbl on
set print demangle on
set demangle-style gnu-v3
</pre><p>
Starting with version 7.0, GDB includes support for writing
pretty-printers in Python. Pretty printers for containers and other
classes are distributed with GCC from version 4.5.0 and should be installed
alongside the libstdc++ shared library files and found automatically by
GDB.
</p><p>
Depending where libstdc++ is installed, GDB might refuse to auto-load
the python printers and print a warning instead.
If this happens the python printers can be enabled by following the
instructions GDB gives for setting your <code class="code">auto-load safe-path</code>
in your <code class="filename">.gdbinit</code> configuration file.
</p><p>
Once loaded, standard library classes that the printers support
should print in a more human-readable format. To print the classes
in the old style, use the <strong class="userinput"><code>/r</code></strong> (raw) switch in the
print command (i.e., <strong class="userinput"><code>print /r foo</code></strong>). This will
print the classes as if the Python pretty-printers were not loaded.
</p><p>
For additional information on STL support and GDB please visit:
<a class="link" href="https://sourceware.org/gdb/wiki/STLSupport" target="_top"> "GDB Support
for STL" </a> in the GDB wiki. Additionally, in-depth
documentation and discussion of the pretty printing feature can be
found in "Pretty Printing" node in the GDB manual. You can find
on-line versions of the GDB user manual in GDB's homepage, at
<a class="link" href="https://sourceware.org/gdb/" target="_top"> "GDB: The GNU Project
Debugger" </a>.
</p></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="debug.req"></a>Debug Versions of Library Binary Files</h3></div></div></div><p>
As described above, libstdc++ is built with debug symbols enabled by default,
but because it's also built with optimizations the code can be hard to
follow when stepping into the library in a debugger.
</p><p>
If you would like to debug <code class="filename">libstdc++.so</code> itself,
there are two ways to build an unoptimized libstdc++ with debug flags.
The first is to create a separate debug build by running make from the
top-level of a tree freshly-configured with
</p><pre class="programlisting">
--enable-libstdcxx-debug
</pre><p>and perhaps</p><pre class="programlisting">
--enable-libstdcxx-debug-flags='...'
</pre><p>
Both the normal build and the debug build will persist, without
having to specify <code class="code">CXXFLAGS</code>, and the debug library will
be installed in a separate directory tree, in <code class="code">(prefix)/lib/debug</code>.
For more information, look at the
<a class="link" href="configure.html" title="Configure">configuration</a> section.
</p><p>
A second approach is to use the configuration flags
</p><pre class="programlisting">
make CXXFLAGS='-g3 -fno-inline -O0' all
</pre></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="debug.compile_time_checks"></a>Compile Time Checking</h3></div></div></div><p> The <a class="link" href="ext_compile_checks.html" title="Chapter 16. Compile Time Checks">Compile-Time
Checks</a> extension has compile-time checks for many algorithms.
These checks were designed for C++98 and have not been updated to work
with C++11 and later standards. They might be removed at a future date.
</p></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="using_exceptions.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="using.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="std_contents.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">Exceptions </td><td width="20%" align="center"><a accesskey="h" href="../index.html">Home</a></td><td width="40%" align="right" valign="top"> Part II.
Standard Contents
</td></tr></table></div></body></html>