Information on known bugs, details on efforts to fix them, and
fixed bugs are all available as part of the GCC bug tracking system,
with the category set to libstdc++
.
Everybody's got issues. Even the C++ Standard Library.
The Library Working Group, or LWG, is the ISO subcommittee responsible for making changes to the library. They periodically publish an Issues List containing problems and possible solutions. As they reach a consensus on proposed solutions, we often incorporate the solution.
Here are the issues which have resulted in code changes to the library. The links are to the specific defect reports from a partial copy of the Issues List. You can read the full version online at the ISO C++ Committee homepage, linked to on the GCC "Readings" page. If you spend a lot of time reading the issues, we recommend downloading the ZIP file and reading them locally.
(NB: partial copy means that not all links within the lwg-*.html pages will work. Specifically, links to defect reports that have not been accorded full DR status will probably break. Rather than trying to mirror the entire issues list on our overworked web server, we recommend you go to the LWG homepage instead.)
If a DR is not listed here, we may simply not have gotten to
it yet; feel free to submit a patch. Search the include/bits
and src directories for appearances of
_GLIBCXX_RESOLVE_LIB_DEFECTS
for examples
of style. Note that we usually do not make changes to the
code until an issue has reached DR status.
This should be two overloaded functions rather than a single function.
Apparently extracting Boolean values was messed up...
If codecvt::do_in
returns noconv
there are
no changes to the values in [to, to_limit)
.
Re-opening a file stream does not clear the state flags.
Implement the proposed resolution.
Padding issues.
An instance of ios_base::failure
is constructed instead.
The return type is the previous state of synchronization.
These members functions are declared private
and are
thus inaccessible. Specifying the correct semantics of
"copying stream state" was deemed too complicated.
This DR made many widespread changes to basic_istream
and basic_ostream
all of which have been implemented.
Make the policy consistent with that of formatted input, unformatted input, and formatted output.
And they do now. An editing glitch in the last item in the list of [27.6.1.2.3]/7.
The text of the standard was gibberish. Typos gone rampant.
Change the first parameter to stateT&
and implement
the new effects paragraph.
Safety checks on the size of the string should test against
max_size()
rather than npos
.
The effect contain isspace(c,getloc())
which must be
replaced by isspace(c,is.getloc())
.
They behave as a formatted input function and as an unformatted
input function, respectively (except that getline
is
not required to set gcount
).
For associative containers where the value type is the same as
the key type, both iterator
and const_iterator
are constant iterators.
The binder1st
and binder2nd
didn't have an
operator()
taking a non-const parameter.
This was not a const member function. Note that the DR says to replace the function with a const one; we have instead provided an overloaded version with identical contents.
num_put::put()
was overloaded on the wrong types.
Same as 117, but for num_get::get()
.
These functions set failbit
on error now.
Make member erase
return iterator for set
, multiset
, map
, multimap
.
seekp
should only set the output stream, and
seekg
should only set the input stream.
op<<
with a const char*
was
calculating an incorrect number of characters to write.
Grow efficiently the internal array object.
Quite complex to summarize...
This function used to take its arguments as reference-to-const, now it copies them (pass by value).
Yes, it can, specifically if EOF is reached while skipping whitespace.
If nothing is extracted into the string, op>>
now
sets failbit
(which can cause an exception, etc., etc.).
Both set
and multiset
were missing
overloaded find, lower_bound, upper_bound, and equal_range functions
for const instances.
For conversion from a floating-point type, str.precision()
is specified in the conversion specification.
Implement N1780, first check before then check after, insert as close to hint as possible.
The declaration of reverse_iterator
lists a default constructor.
However, no specification is given what this constructor should do.
Add a helper for forward_iterator/output_iterator, fix the existing one for input_iterator/output_iterator to not rely on Assignability.
Store a null character only if the character array has a non-zero size.
This nested typedef was originally not specified.
Make the copy constructor and copy-assignment operator declarations public in gslice_array, indirect_array, mask_array, slice_array; provide definitions.
The default ctor would build its members from copies of temporaries; now it simply uses their respective default ctors.
The bad_
* classes no longer have destructors (they
are trivial), since no description of them was ever given.
The typedefs it inherits from its base classes can't be used, since
(for example) basic_iostream<T>::traits_type
is ambiguous.
Similar to 118.
Add global functions with two template parameters. (NB: not added for now a templated assignment operator)
If (this == &rhs)
do nothing.
If (this == &x)
do nothing.
Basically, compare the input character to
is.widen(0)
and is.widen(1)
.
Do not specify what codecvt<wchar_t, char,
mbstate_t>::do_length
must return.
Change the format string to "%.0Lf".
Add const overloads of is_open
.
Add the real(T)
and imag(T)
members; in C++11 mode, also adjust the existing
real()
and imag()
members and
free functions.
Change it to return a const T&
.
Implement the proposed resolution.
Replace "new" with "::new".
Tweak the debug-mode checks in _Safe_iterator.
Have open
clear the error flags.
Implement Option 3, as per N1599.
Implement the resolution, beyond DR 169.
Add three overloads, taking fewer template arguments.
Implement the resolution, basically cast less.
Don't fail if the next pointer is null and newoff is zero.
Initialize cerr tied to cout and wcerr tied to wcout.
Add data()
to std::vector
and
at(const key_type&)
to std::map
.
Fix the parameters.
Construct a linear_congruential
engine and seed with it.
Use &value.
In case of input_iterator/output_iterator rely on Assignability of input_iterator' value_type.
We were almost doing the right thing, just use std::move in adjacent_difference.
Add an auto_ptr<void> specialization.
Follow the straightforward proposed resolution.
In C++11 mode, remove the pow(float,int), etc., signatures.
Change it to be a formatted output function (i.e. catch exceptions).
Add the missing modes to fopen_mode.
Implement the simple resolution.
Add the missing operations.
In C++11 mode add cbegin(size_type) and cend(size_type) to the unordered containers.
Add it, consistently with the discussion.
Make the member functions table and classic_table public.
Implement the straightforward resolution.
In C++11 mode, add at() and at() const.
Implement the int -> size_t replacements.
In C++11 mode, remove assign, add fill.
In C++11 mode, add std::proj.
Add the overload.
In C++11 mode, remove the pow(complex<T>, int) signature.
Update / add the signatures.
The traditional HP / SGI return type and value is blessed by the resolution of the DR.