Turning off explicit template instantiation leads to a pretty
significant build time and code size cost. We're better off dealing
with ABI incompatibility issues that come up in a less heavy handed
way.
This reverts commit r189610.
llvm-svn: 215740
libc++ currently relies on undefined initialization order of global
initializers when using gcc:
1. __start_std_streams in iostream.cpp calls locale:🆔:_init, which assigns
an id to each locale::facet in an initializer
2. Every facet has a static locale::id id, whose constructor sets the facet's
id to 0
If 2 runs after 1, it clobbers the facet's assigned consecutive id, causing
exceptions to be thrown when e.g. running code like "cout << endl".
To fix this, let _LIBCPP_CONSTEXPR evaluate to "constexpr" instead of nothing
with gcc. locale::id's constructor is marked _LIBCPP_CONSTEXPR, which ensures
that it won't get an initializer that could potentially run after the
iostream.cpp initializer. (This remains broken when building with msvc.)
Also switch constexpr-specific code in bitset to use __SIZEOF_SIZE_T__ instead
of __SIZE_WIDTH__, because gcc doesn't define the latter.
Pair-programmed/debugged with Dana Jansens.
llvm-svn: 210188
libc++ will not call address_sanitizer to detect addressing errors in the
standard library containers. This is a negative macro to enable users to
disable the libc++ checks even if they are compiling with address sanitizer
enabled by defining this macro.
At the present time, there is no code in libc++ that looks at this macro.
That will come soon. This is just infrastructure.
llvm-svn: 206184
ARM64 generates RTTI with hidden visibility, which means that typeinfo
must be compared char-by-char since it's not guaranteed to be uniqued
across the whole program.
llvm-svn: 205139
with Visual C++ 2013 by making libcxx definition text-identical to yvals.h.
Persumably this definition is for older Visual C++ versions.
In such cases it will still be defined so no functionality change.
Other platforms should not be affected as this is inside
#elif defined(_LIBCPP_MSVC)
Patch by G M!
llvm-svn: 201328
GCC does not support strong enum if -std=c++0x is not used.
Without the strong enum, we will see following error:
In file included from libcxx/include/ostream:131:0,
from libcxx/include/sstream:174,
from libcxx/include/complex:247,
from cpp03-headers.cpp:11:
libcxx/include/ios:419:68: error: 'io_errc' is not a class or namespace
libcxx/include/ios:420:66: error: 'io_errc' is not a class or namespace
To workaround this issue, this commit will define
_LIBCPP_HAS_NO_STRONG_ENUMS when we are compiling with
g++ without c++0x.
llvm-svn: 197313
type_traits:3280:31: error: expected primary-expression before 'decltype'
type_traits:3280:29: error: expected ';' at end of member declaration
memory:2415:49: error: function 'std::__1::default_delete<_Tp>::default_delete()'
defaulted on its first declaration must not have an exception-specification
memory:2435:49: error: function 'std::__1::default_delete<_Tp []>::default_delete()'
defaulted on its first declaration must not have an exception-specification
The attached patch defines _LIBCPP_HAS_NO_ADVANCED_SFINAE and
_LIBCPP_HAS_NO_DEFAULTED_FUNCTIONS for gcc version < 4.7, making
the library compile with gcc 4.6.4.
llvm-svn: 195431
trivial in C++03, thus making it trivial in both C++03 and C++11.
This patch allows one to opt-in/out of this decision with a macro. You can
choose to have the pair copy constructor always be trivial, or always be
non-trivial. The flag controlling this is now _LIBCPP_TRIVIAL_PAIR_COPY_CTOR.
The client can define this flag to 1, and the pair copy constructor will be
trivial (when possible of course), or to 0, and the pair copy constructor will
be nontrivial.
Default settings for this flag are set in <__config> (as usual). With this
commit the default is _LIBCPP_TRIVIAL_PAIR_COPY_CTOR=1 for all platforms
except __APPLE__, which defaults to _LIBCPP_TRIVIAL_PAIR_COPY_CTOR=0.
llvm-svn: 194742
pair, and a couple of pair-like implementation detail types. The
C++98/03 and 11 standards all specify that the copy constructor of
pair<int, int> is trivial. However as libc++ tracked the draft C++11
standard over the years, this copy constructor became non-trivial, and
then just recently was corrected back to trivial for C++11.
Unfortunately (for libc++1) the Itanium ABI specifies different calling
conventions for trivial and non-trivial copy constructors. Therefore
currently the C++03 libc++ copy constructor for pair<int, int> is ABI
incompatible with the C++11 libc++ copy constructor for pair<int, int>.
This is Bad(tm). This patch corrects the situation by making this copy
constructor trivial in C++03 mode as well.
Just in case it is needed for an incomplete C++11 compiler, libc++
retains the ability to support pair with rvalue references, but without
defaulted special members. However the pair needs non-trivial special
members to implement this special case, (as it did when clang was in
this place a couple of years ago).
During this work a bug was also found and fixed in
is_trivially_constructible.
And there is a minor drive-by fix in <__config> regarding
__type_visibility__.
A test is updated to ensure that the copy constructor of pair<int, int>
is trivial in both C++03 and C++11. This test will necessarily fail for
a compiler that implements rvalue references but not defaulted special
members.
llvm-svn: 194536
The issue this patch seeks to address is that MS's compiler (cl.exe) doesn't support the __attribute__((__weak__)) or __atribute__((__visibility__("default")) syntax; so a solution must be found where cl.exe doesn't see this syntax.
This patch seeks to solve this problem by changing code patterned like this:
__attribute__((__weak__, __visibility__("default")))
void* operator new(size_t size, const std::nothrow_t&) _NOEXCEPT { /*snip*/; return p; }
to code like this:
_LIBCPP_WEAK
void* operator new(size_t size, const std::nothrow_t&) _NOEXCEPT { return p; }
Howard: Thanks for all the comments regarding the default visibility
tag on the definition. I agree it isn't needed, and that there are lots
of other places where it is missing. That being said, I'm not wanting
to rock the boat on that issue right now. So I've added it back to the
definition via _LIBCPP_FUNC_VIS. A later pass dedicated just to this
issue can bring things in to a consistent state one way or the other.
Note that we do not want to have the exact same attributes on the
declaration and defintion in this case. The declaration should not be
marked weak, whereas the definition should (which is what G M's patch
did). I've fully tested on OS X to ensure that the resultant attribute
syntax actually works.
llvm-svn: 192007
1. I had been detecting and trapping iterator == and \!= among iterators
in different containers as an error. But the trapping itself is actually
an error.
Consider:
#include <iostream>
#include <vector>
#include <algorithm>
template <class C>
void
display(const C& c)
{
std::cout << "{";
bool first = true;
for (const auto& x : c)
{
if (\!first)
std::cout << ", ";
first = false;
std::cout << x;
}
std::cout << "}\n";
}
int
main()
{
typedef std::vector<int> V;
V v1 = {1, 3, 5};
V v2 = {2, 4, 6};
display(v1);
display(v2);
V::iterator i = std::find(v1.begin(), v1.end(), 1);
V::iterator j = std::find(v2.begin(), v2.end(), 2);
if (*i == *j)
i = j; // perfectly legal
// ...
if (i \!= j) // the only way to check
v2.push_back(*i);
display(v1);
display(v2);
}
It is legal to assign an iterator from one container to another of the
same type. This is required to work. One might want to test whether or
not such an assignment had been made. The way one performs such a check
is using the iterator's ==, \!= operator. This is a logical and necessary
function and does not constitute an error.
2. I had a header circular dependence bug when _LIBCPP_DEBUG2 is defined.
This caused a problem in several of the libc++ tests.
Fixed.
3. There is a serious problem when _LIBCPP_DEBUG2=1 at the moment in that
std::basic_string is inoperable. std::basic_string uses __wrap_iterator
to implement its iterators. __wrap_iterator has been rigged up in debug
mode to support vector. But string hasn't been rigged up yet. This means
that one gets false positives when using std::string in debug mode. I've
upped std::string's priority in www/debug_mode.html.
llvm-svn: 187636
Viktor Kutuzov