libstdc++ Frequently Asked Questions
The latest version of this document is always available at
http://gcc.gnu.org/onlinedocs/libstdc++/faq/.
To the libstdc++-v3 homepage.
Questions
- General Information
- What is libstdc++-v3?
- Why should I use libstdc++?
- Who's in charge of it?
- How do I get libstdc++?
- When is libstdc++ going to be finished?
- How do I contribute to the effort?
- What happened to libg++? I need that!
- What if I have more questions?
- Installation
- How do I install libstdc++-v3?
- Is this a drop-in replacement for the
libstdc++ that's shipped with g++?
- What is this CVS thing that you keep
mentioning?
- How do I know if it works?
- Platform-Specific Issues
- Can libstdc++-v3 be used with <my
favorite compiler>?
- Building under Cygwin hangs/explodes!?
- Known Bugs and Non-Bugs
- What works already?
- Bugs in gcc/g++ (not libstdc++-v3)
- Bugs in the C++ language/lib specification
- Things in libstdc++ that look like bugs
The g++-3 headers are
not ours
- Aw, that's easy to fix!
- Miscellaneous
- string::iterator is not char*;
vector<T>::iterator is not T*
- What's next after libstdc++-v3?
- What about the STL from SGI?
- Extensions and Backward Compatibility
- Compiling with "-fnew-abi"
- Is libstdc++-v3 thread-safe?
- How do I get a copy of the ISO C++ Standard?
The GNU Standard C++ Library v3, or libstdc++-2.90.x/2.9x, is an
ongoing project to implement the ISO 14882 Standard C++ library
as described in chapters 17 through 27 and annex D. As the
library reaches stable plateaus, it is captured in a snapshot
and released. The current release is the
tenth snapshot. For those who want to see exactly how
far the project has come, or just want the latest
bleeding-edge code, the up-to-date source is available over
anonymous CVS, and can even be browsed over the Web (see below).
A more formal description of the V3 goals can be found in the
official design document.
The completion of the ISO C++ standardization gave the
C++ community a powerful set of reuseable tools in the form
of the C++ Standard Library. However, all existing C++
implementations are (as the Draft Standard used to say)
"incomplet and incorrekt," and many suffer from
limitations of the compilers that use them.
The GNU C/C++/FORTRAN/<pick-a-language> compiler
(gcc, g++, etc) is widely considered to be
one of the leading compilers in the world. Its development
has recently been taken over by the
GCC team. All of
the rapid development and near-legendary
portability
that are the hallmarks of an open-source project are being
applied to libstdc++.
That means that all of the Standard classes and functions
(such as string, vector<>, iostreams,
and algorithms) will be freely available and fully compliant.
Programmers will no longer need to "roll their own"
nor be worried about platform-specific incompatabilities.
The libstdc++ project is contributed to by several developers
all over the world, in the same way as GCC or Linux.
Benjamin Kosnik, Gabriel Dos Reis, Phil Edwards, and Ulrich
Drepper are the lead maintainers of the CVS archive.
Development and discussion is held on the libstdc++ mailing
list. Subscribing to the list, or searching the list
archives, is open to everyone. You can read instructions for
doing so on the homepage. If you
have questions, ideas, code, or are just curious, sign up!
The tenth (and latest) snapshot of libstdc++-v3 is
available via ftp.
The homepage
has instructions for retrieving the latest CVS sources, and for
browsing the CVS sources over the web.
The subset commonly known as the Standard Template Library
(chapters 23 through 25, mostly) is adapted from the SGI STL,
which is also an ongoing work.
Nathan Myers gave the best of all possible answers, responding
to a Usenet article asking this question: Sooner, if you help.
Here is a
page devoted to this topic. Subscribing to the mailing
list (see above, or the homepage) is a very good idea if you
have something to contribute, or if you have spare time and
want to help. Contributions don't have to be in the form of
source code; anybody who is willing to help write
documentation, for example, or has found a bug in code that
we all thought was working, is more than welcome!
The most recent libg++ README states that libg++ is no longer
being actively maintained. It should not be used for new
projects, and is only being kicked along to support older code.
The libg++ was designed and created when there was no Standard
to provide guidance. Classes like linked lists are now provided
for by list<T> and do not need to be created by
genclass. (For that matter, templates exist now and
are well-supported, whereas genclass (mostly) predates them.)
There are other classes in libg++ that are not specified in the
ISO Standard (e.g., statistical analysis). While there are a
lot of really useful things that are used by a lot of people
(e.g., statistics :-), the Standards Committee couldn't include
everything, and so a lot of those "obvious" classes
didn't get included.
Since libstdc++ is an implementation of the Standard Library, we
have no plans at this time to include non-Standard utilities
in the implementation, however handy they are. (The extensions
provided in the SGI STL aren't maintained by us and don't get
a lot of our attention, because they don't require a lot of our
time.) It is entirely plausable that the "useful stuff"
from libg++ might be extracted into an updated utilities library,
but nobody has stated such a project yet.
(The Boost site houses free
C++ libraries that do varying things, and happened to be started
by members of the Standards Committee. Certain "useful
stuff" classes will probably migrate there.)
For the bold and/or desperate, the
GCC FAQ
describes where to find the last libg++ source.
If you have read the README and RELEASE-NOTES files, and your
question remains unanswered, then just ask the mailing list.
At present, you do not need to be subscribed to the list to
send a message to it. More information is available on the
homepage (including how to browse the list archives); to send
to the list, use
libstdc++@gcc.gnu.org.
If you have a question that you think should be included here,
or if you have a question about a question/answer here,
contact Phil Edwards
or Gabriel Dos Reis.
Complete instructions are not given here (this is a FAQ, not
an installation document), but the tools required are few:
- A release of libstdc++.
- A recent release of GCC (version 2.95 works). Note
that building GCC is much easier and more automated
than building the GCC 2.[78] series was.
- If you plan on hacking around with the makefiles, you
will need the tools autoconfand automake.
- GNU Make is the only make that supports these makefiles.
The file documentation.html
provides a good overview of the steps necessary to build, install,
and use the library. Instructions for configuring the library
with new flags such as --enable-threads are there also.
The top-level install.html and
RELEASE-NOTES files contain
the exact build and installation instructions. You may wish to
browse those files over CVSweb ahead of time to get a feel for
what's required. RELEASE-NOTES is located in the
".../docs/17_intro/" directory of the distribution.
Yes, as of 2.90.8, it is intended as such. And as of 2.91,
libstdc++-v3 is the library that's shipped with
g++, so much of this answer has become moot.
The Concurrent Versions System is one of several revision
control packages. It was selected for GNU projects because it's
free (speech), free (beer), and very high quality. The CVS entry in
the GNU software catalogue has a better description as
well as a
link to the makers of CVS.
The "anonymous client checkout" feature of CVS is
similar to anonymous FTP in that it allows anyone to retrieve
the latest libstdc++ sources.
After the first of April, American users will have a
"/pharmacy" command-line option...
libstdc++-v3 comes with its own testsuite. You do not need
to actually install the library ("gmake
install") to run the testsuite. Note that 2.91 does
not use DejaGNU yet.
To run the testsuite on the library after building it, use
"gmake check" while in your build directory. To run
the testsuite on the library after building and installing it,
use "gmake check-install" instead.
The testsuite subdirectory in your build directory will then
contain three files of the form YYYYMMDD-mkcheck*.txt. One of
them (-mkcheck.txt itself) contains the results of the tests;
this can be mailed to the list. The other files (-mkchecklog.txt
and -mkcheckfiles.txt) contain messages from the compiler while
building the test programs, and a list of the tests to be run,
respectively.
If you find bugs in the testsuite programs themselves, or if
you think of a new test program that should be added to the
suite, please write up your idea and send it to the list!
Probably not. Yet.
Because GCC advances so rapidly, development and testing of
libstdc++ is being done almost entirely under that compiler.
If you are curious about whether other, lesser compilers
(*grin*) support libstdc++, you are more than welcome to try.
Configuring and building the library (see above) will still
require certain tools, however. Also keep in mind that
building libstdc++ does not imply that your compiler
will be able to use all of the features found in the
C++ Standard Library.
Since the goal of ISO Standardization is for all C++
implementations to be able to share code, the final libstdc++
should, in theory, be useable under any ISO-compliant
compiler. It will still be targeted and optimized for
GCC/g++, however.
Sometimes, yes. You're probably in the middle of generating
the numeric_limits specializations when it hangs,
right? Thought so...
The <limits> header and its associated library
code are platform-specific. These files get generated from
scratch during installation, and it is this generator that is
hanging. More specifically, the only sure way to determine
what the numeric_limits<T>::traps boolean
should be is to actually divide by zero and see if it is
trapped or not.
Under NT, this will occasionally just hang. On those
occasions when the test does not hang, the zero-division is
in fact trapped. That doesn't prevent hanging elsewhere.
You have two options. You can get a newer cygwin1.dll (see the
Cygwin paragraph in the
installation instructions).
Or you can get a prebuilt set of bits/std_limits.h and
src/limitsMEMBERS.cc files from Mumit Khan's
Cygwin-related website.
Note that this section can get rapdily outdated -- such is the
nature of an open-source project. For the latest information, join
the mailing list or look through recent archives. The RELEASE-
NOTES and BUGS files are generally kept up-to-date.
This is a verbatim clip from the "Status" section
of the RELEASE-NOTES for the latest snapshot.
New:
- namespace std:: is now on by default.
- choice of "C" include strategies, including the shadow header work,
or generic global to std mapping of required "C" types.
- cpu/atomicity.h tweaks, additions of ia64 and arm support.
- abstraction of atomicity.h header to support notion of os/atomicity.h files.
- addition of backward header bits
- use of system_header pragma
- Conditional use of -Werror
- preliminary support for new g++ diagnostics capabilities, including
-fdiagnostics-show-location=once
- pedantic and shadow argument warning fixes
- Ugly, yet correct mechanism for dealing with "C" math adopted,
including the use of builtins.
- updates and configure/build work to support new libtool
- addition of strstream
- valarray work
- complex work
- update to SGI STL 3.3
- libio sync between glibc/libstdc++-v3. Some divergence since initial
merge, but sources remain quite close.
- IO fixes for alpha
- wide character work for IO when using libio
- addition of c_io_stdio and "C" IO abstraction layer.
- auto_ptr fixes, testsuite additions
- Attempts to use -ffunction-sections -fdata-sections and
--gc-sections, depending on use of GNU ld and specific features. As of
late, --gc-sections has been disabled due to problems with it throwing
away initialization sections. This work is ongoing.
- long double support
- sub directory removal, coherent organization of cpu and os-specific
files, consolidation of include directories, integration of the C++
support bits for operator new/delete,exceptions, etc. All includes
are now either in the include/* hierarchy or in libsupc++'s sub directory.
- Support for more platforms, including irix and bsd variants.
- filebuf tweaks to deal with variable-size buffers.
- filebuf implementation for putbackc, etc. al.
- ctype rewritten. Includes ctype, ctype, and others.
- codecvt rewritten. Includes codecvt,
codecvt. In addition,
implementation-defined conversions using iconv are now supported with
the __enc_traits partial-specialization of the State template
parameter of the codecvt class. In this manner, conversions between
encodings such as UCS4, USC2, UNICODE, UNICODEBIG, UNICODELITTLE, etc
can be performed.
- preliminary work on named locales
- preliminary documentation for locale implementation has been established.
- Many, many bug fixes.
- Many, many testsuite additions and consistent VERIFY usage.
- work on mkcheck to make it more flexible, use libtool, etc.
This is by no means meant to be complete nor exhaustive, but
mentions some problems that users may encounter when building
or using libstdc++. If you are experiencing one of these
problems, you can find more information on the libstdc++ and
the GCC mailing lists.
- As of 2.91, these bugs have all been fixed. We look forward
to new ones, well, not exactly...
Yes, unfortunately, there are some. In a message
to the list, Nathan Myers announced that he has started a list of
problems in the ISO C++ Standard itself, especially with
regard to the chapters that concern the library. The list
itself is posted on his
website. Developers who are having problems interpreting
the Standard may wish to consult his notes.
For those people who are not part of the ISO Library Group
(i.e., nearly all of us needing to read this page in the first
place :-), a public list of the library defects is occasionally
published here.
There are things which are not bugs in the compiler (4.2) nor
the language specification (4.3), but aren't really bugs in
libstdc++, either. Really!
The biggest of these is the quadzillions of warnings about the
library headers emitted when -Weffc++ is used. Making
libstdc++ "-Weffc++-clean" is not a goal of the project,
for a few reasons. Mainly, that option tries to enforce
object-oriented programming, while the Standard Library isn't
necessarily trying to be OO. There are multiple solutions
under discussion.
If you have found an extremely broken header file which is
causing problems for you, look carefully before submitting a
"high" priority bug report (which you probably shouldn't
do anyhow; see the last paragraph of the page describing
the GCC bug database).
If the headers are in ${prefix}/include/g++-3, then
you are using the old libstdc++-v2 library, which is nonstandard
and unmaintained. Do not report problems with -v2 to the -v3
mailing list.
Currently our header files are installed in
${prefix}/include/g++-v3 (see the 'v'?). This may
change with the next release of GCC, as it may be too confusing,
but the
question has not yet been decided.
If you have found a bug in the library and you think you have
a working fix, then send it in! The main GCC site has a page
on submitting
patches that covers the procedure, but for libstdc++ you
should of course send the patch to our mailing list, not the
GCC mailing list. The libstdc++
contributors' page
also talks about how to submit patches.
In addition to the description, the patch, and the ChangeLog
entry, it is a Good Thing if you can additionally create a small
test program to test for the presence of the bug that your
patch fixes. Bugs have a way of being reintroduced; if an old
bug creeps back in, it will be caught immediately by the
testsuite -- but only if such a test exists.
If you have code that depends on container<T> iterators
being implemented as pointer-to-T, your code is broken.
While there are arguments for iterators to be implemented in
that manner, A) they aren't very good ones in the long term,
and B) they were never guaranteed by the Standard anyway. The
type-safety achieved by making iterators a real class rather
than a typedef for T* outweighs nearly all opposing
arguments.
Hopefully, not much. The goal of libstdc++-v3 is to produce
a fully-compliant, fully-portable Standard Library. After that,
we're mostly done: there won't be any more compliance
work to do.
The ISO Committee will meet periodically to review Defect Reports
in the C++ Standard. Undoubtably some of these will result in
changes to the Standard, which will be reflected in patches to
libstdc++. Some of that is already happening, see 4.2. Some of
those changes are being predicted by the library maintainers, and
we add code to the library based on what the current proposed
resolution specifies.
The current libstdc++ contains extensions to the Library which
must be explicitly requested by client code (for example, the
hash tables from SGI). Other extensions may be added to
libstdc++-v3 if they seem to be "standard" enough.
(For example, the "long long" type from C99.)
Bugfixes and rewrites (to improve or fix thread safety, for
instance) will of course be a continuing task.
This
question about the next libstdc++ prompted some brief but
interesting speculation.
The STL from SGI
is merged into libstdc++-v3 with changes as necessary.
Currently release 3.3 is being used. Changes in the STL
usually produce some weird bugs and lots of changes in the
rest of the libstd++ source as we scramble to keep up. :-)
In particular, string is not from SGI and makes no
use of their "rope" class (which is included as an
optional extension), nor is valarray and some others.
Classes like vector<> are, however.
The FAQ for SGI's STL (one jump off of their main page) is
recommended reading.
Although you can specify -I options to make the
preprocessor search the g++-v3/ext and /backward directories,
it is better to refer to files there by their path, as in:
#include <ext/hash_map>
Extensions to the library have
their own page.
Towards the end of July 1999, this subject was brought up again
on the mailing list under a different name. The related
thread
(by the name HOWTO-honor-std) is very instructive. More info
is at the end of RELEASE-NOTES.
This functionality is now automated and turned on by default.
Quick answer: no, as of 2.91 (tenth snapshot), the
library is not appropriate for multithreaded access. The
string class is MT-safe.
This is assuming that your idea of "multithreaded"
is the same as ours... The general question of multithreading
and libstdc++-v3 is addressed in the chapter-specific advice for
Library
Introduction. Threadsafe containers are covered in
more detail in
the
Received Wisdom section on containers.
Copies of the full ISO 14882 standard are available on line
via the ISO mirror site for committee members. Non-members,
or those who have not paid for the privilege of sitting on
the committee and sustained their two-meeting commitment for
voting rights, may get a copy of the standard from their
respective national standards organization. In the USA,
this national standards organization is ANSI and their
website is right here.
(And if you've already registered with them, clicking this
link will take you to directly to the place where you can
buy
the standard on-line.
Who is your country's member body? Visit the
ISO homepage and find out!
Comments and suggestions are welcome, and may be sent to
Phil Edwards or
Gabriel Dos Reis.
$Id: index.html,v 1.10 2000/12/03 23:47:49 jsm28 Exp $