maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/libcxx/include/istream

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1660 lines
47 KiB
Plaintext
Raw Normal View History

libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
// -*- C++ -*-
//===--------------------------- istream ----------------------------------===//
//
Update more file headers across all of the LLVM projects in the monorepo to reflect the new license. These used slightly different spellings that defeated my regular expressions. We understand that people may be surprised that we're moving the header entirely to discuss the new license. We checked this carefully with the Foundation's lawyer and we believe this is the correct approach. Essentially, all code in the project is now made available by the LLVM project under our new license, so you will see that the license headers include that license only. Some of our contributors have contributed code under our old license, and accordingly, we have retained a copy of our old license notice in the top-level files in each project and repository. llvm-svn: 351648
2019-01-19 18:56:40 +08:00
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
//
//===----------------------------------------------------------------------===//
#ifndef _LIBCPP_ISTREAM
#define _LIBCPP_ISTREAM
/*
istream synopsis
template <class charT, class traits = char_traits<charT> >
class basic_istream
: virtual public basic_ios<charT,traits>
{
public:
// types (inherited from basic_ios (27.5.4)):
typedef charT char_type;
typedef traits traits_type;
typedef typename traits_type::int_type int_type;
typedef typename traits_type::pos_type pos_type;
typedef typename traits_type::off_type off_type;
// 27.7.1.1.1 Constructor/destructor:
explicit basic_istream(basic_streambuf<char_type, traits_type>* sb);
basic_istream(basic_istream&& rhs);
virtual ~basic_istream();
// 27.7.1.1.2 Assign/swap:
basic_istream& operator=(basic_istream&& rhs);
void swap(basic_istream& rhs);
// 27.7.1.1.3 Prefix/suffix:
class sentry;
// 27.7.1.2 Formatted input:
basic_istream& operator>>(basic_istream& (*pf)(basic_istream&));
basic_istream& operator>>(basic_ios<char_type, traits_type>&
(*pf)(basic_ios<char_type, traits_type>&));
basic_istream& operator>>(ios_base& (*pf)(ios_base&));
basic_istream& operator>>(basic_streambuf<char_type, traits_type>* sb);
basic_istream& operator>>(bool& n);
basic_istream& operator>>(short& n);
basic_istream& operator>>(unsigned short& n);
basic_istream& operator>>(int& n);
basic_istream& operator>>(unsigned int& n);
basic_istream& operator>>(long& n);
basic_istream& operator>>(unsigned long& n);
basic_istream& operator>>(long long& n);
basic_istream& operator>>(unsigned long long& n);
basic_istream& operator>>(float& f);
basic_istream& operator>>(double& f);
basic_istream& operator>>(long double& f);
basic_istream& operator>>(void*& p);
// 27.7.1.3 Unformatted input:
streamsize gcount() const;
int_type get();
basic_istream& get(char_type& c);
basic_istream& get(char_type* s, streamsize n);
basic_istream& get(char_type* s, streamsize n, char_type delim);
basic_istream& get(basic_streambuf<char_type,traits_type>& sb);
basic_istream& get(basic_streambuf<char_type,traits_type>& sb, char_type delim);
basic_istream& getline(char_type* s, streamsize n);
basic_istream& getline(char_type* s, streamsize n, char_type delim);
basic_istream& ignore(streamsize n = 1, int_type delim = traits_type::eof());
int_type peek();
basic_istream& read (char_type* s, streamsize n);
streamsize readsome(char_type* s, streamsize n);
basic_istream& putback(char_type c);
basic_istream& unget();
int sync();
pos_type tellg();
basic_istream& seekg(pos_type);
basic_istream& seekg(off_type, ios_base::seekdir);
Some of the synopsis was left out of these headers, and the copy construction/assignment should have been marked as deleted. Done. No functionality change, because the base class (base_ios) was marked as non-copyable already. llvm-svn: 217876
2014-09-16 23:27:01 +08:00
protected:
basic_istream(const basic_istream& rhs) = delete;
basic_istream(basic_istream&& rhs);
// 27.7.2.1.2 Assign/swap:
basic_istream& operator=(const basic_istream& rhs) = delete;
basic_istream& operator=(basic_istream&& rhs);
void swap(basic_istream& rhs);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
};
// 27.7.1.2.3 character extraction templates:
template<class charT, class traits>
basic_istream<charT,traits>& operator>>(basic_istream<charT,traits>&, charT&);
template<class traits>
basic_istream<char,traits>& operator>>(basic_istream<char,traits>&, unsigned char&);
template<class traits>
basic_istream<char,traits>& operator>>(basic_istream<char,traits>&, signed char&);
template<class charT, class traits>
basic_istream<charT,traits>& operator>>(basic_istream<charT,traits>&, charT*);
template<class traits>
basic_istream<char,traits>& operator>>(basic_istream<char,traits>&, unsigned char*);
template<class traits>
basic_istream<char,traits>& operator>>(basic_istream<char,traits>&, signed char*);
template <class charT, class traits>
void
swap(basic_istream<charT, traits>& x, basic_istream<charT, traits>& y);
typedef basic_istream<char> istream;
typedef basic_istream<wchar_t> wistream;
template <class charT, class traits = char_traits<charT> >
class basic_iostream :
public basic_istream<charT,traits>,
public basic_ostream<charT,traits>
{
public:
// types:
typedef charT char_type;
typedef traits traits_type;
typedef typename traits_type::int_type int_type;
typedef typename traits_type::pos_type pos_type;
typedef typename traits_type::off_type off_type;
// constructor/destructor
explicit basic_iostream(basic_streambuf<char_type, traits_type>* sb);
basic_iostream(basic_iostream&& rhs);
virtual ~basic_iostream();
// assign/swap
basic_iostream& operator=(basic_iostream&& rhs);
void swap(basic_iostream& rhs);
};
template <class charT, class traits>
void
swap(basic_iostream<charT, traits>& x, basic_iostream<charT, traits>& y);
typedef basic_iostream<char> iostream;
typedef basic_iostream<wchar_t> wiostream;
template <class charT, class traits>
basic_istream<charT,traits>&
ws(basic_istream<charT,traits>& is);
[libc++] Re-apply fdc41e11f (LWG1203) without breaking the C++11 build fdc41e11f was reverted in e46c1def5 because it broke the C++11 build. We shouldn't be using enable_if_t in C++11, instead we must use enable_if<...>::type.
2020-09-23 20:49:00 +08:00
// rvalue stream extraction
template <class Stream, class T>
Stream&& operator>>(Stream&& is, T&& x);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
} // std
*/
#include <__config>
Second part of P0482 - char8_t. Reviewed as https://reviews.llvm.org/D55308 llvm-svn: 348828
2018-12-11 12:35:44 +08:00
#include <version>
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
#include <ostream>
Add protection from min/max macros llvm-svn: 145407
2011-11-30 00:45:27 +08:00
Windows support by Ruben Van Boxem. llvm-svn: 142235
2011-10-18 04:05:10 +08:00
#if !defined(_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
#pragma GCC system_header
Windows support by Ruben Van Boxem. llvm-svn: 142235
2011-10-18 04:05:10 +08:00
#endif
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
[Libc++] Use #pragma push_macro/pop_macro to better handle min/max on Windows Summary: This patch improves how libc++ handles min/max macros within the headers. Previously libc++ would undef them and emit a warning. This patch changes libc++ to use `#pragma push_macro` to save the macro before undefining it, and `#pragma pop_macro` to restore the macros and the end of the header. Reviewers: mclow.lists, bcraig, compnerd, EricWF Reviewed By: EricWF Subscribers: cfe-commits, krytarowski Differential Revision: https://reviews.llvm.org/D33080 llvm-svn: 304357
2017-06-01 06:07:49 +08:00
_LIBCPP_PUSH_MACROS
#include <__undef_macros>
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
_LIBCPP_BEGIN_NAMESPACE_STD
template <class _CharT, class _Traits>
[NFC] Rename _LIBCPP_TYPE_VIS_ONLY to _LIBCPP_TEMPLATE_VIS The name _LIBCPP_TYPE_VIS_ONLY is no longer accurate because both _LIBCPP_TYPE_VIS and _LIBCPP_TYPE_VIS_ONLY expand to __attribute__((__type_visibility__)) with Clang. The only remaining difference is that _LIBCPP_TYPE_VIS_ONLY can be applied to templates whereas _LIBCPP_TYPE_VIS cannot (due to dllimport/dllexport not being allowed on templates). This patch renames _LIBCPP_TYPE_VIS_ONLY to _LIBCPP_TEMPLATE_VIS. llvm-svn: 291035
2017-01-05 07:56:00 +08:00
class _LIBCPP_TEMPLATE_VIS basic_istream
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
: virtual public basic_ios<_CharT, _Traits>
{
streamsize __gc_;
public:
// types (inherited from basic_ios (27.5.4)):
typedef _CharT char_type;
typedef _Traits traits_type;
typedef typename traits_type::int_type int_type;
typedef typename traits_type::pos_type pos_type;
typedef typename traits_type::off_type off_type;
// 27.7.1.1.1 Constructor/destructor:
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
explicit basic_istream(basic_streambuf<char_type, traits_type>* __sb) : __gc_(0)
{ this->init(__sb); }
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
virtual ~basic_istream();
protected:
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#ifndef _LIBCPP_CXX03_LANG
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
inline _LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream(basic_istream&& __rhs);
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
// 27.7.1.1.2 Assign/swap:
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
inline _LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream& operator=(basic_istream&& __rhs);
#endif
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
void swap(basic_istream& __rhs) {
_VSTD::swap(__gc_, __rhs.__gc_);
basic_ios<char_type, traits_type>::swap(__rhs);
}
Some of the synopsis was left out of these headers, and the copy construction/assignment should have been marked as deleted. Done. No functionality change, because the base class (base_ios) was marked as non-copyable already. llvm-svn: 217876
2014-09-16 23:27:01 +08:00
Replace _LIBCPP_HAS_NO_DELETED_FUNCTIONS with _LIBCPP_CXX03_LANG llvm-svn: 291278
2017-01-07 04:58:25 +08:00
#ifndef _LIBCPP_CXX03_LANG
Forgot 'const' on my last checkin llvm-svn: 217877
2014-09-16 23:33:53 +08:00
basic_istream (const basic_istream& __rhs) = delete;
basic_istream& operator=(const basic_istream& __rhs) = delete;
Some of the synopsis was left out of these headers, and the copy construction/assignment should have been marked as deleted. Done. No functionality change, because the base class (base_ios) was marked as non-copyable already. llvm-svn: 217876
2014-09-16 23:27:01 +08:00
#endif
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
public:
// 27.7.1.1.3 Prefix/suffix:
[NFC] Rename _LIBCPP_TYPE_VIS_ONLY to _LIBCPP_TEMPLATE_VIS The name _LIBCPP_TYPE_VIS_ONLY is no longer accurate because both _LIBCPP_TYPE_VIS and _LIBCPP_TYPE_VIS_ONLY expand to __attribute__((__type_visibility__)) with Clang. The only remaining difference is that _LIBCPP_TYPE_VIS_ONLY can be applied to templates whereas _LIBCPP_TYPE_VIS cannot (due to dllimport/dllexport not being allowed on templates). This patch renames _LIBCPP_TYPE_VIS_ONLY to _LIBCPP_TEMPLATE_VIS. llvm-svn: 291035
2017-01-05 07:56:00 +08:00
class _LIBCPP_TEMPLATE_VIS sentry;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
// 27.7.1.2 Formatted input:
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
basic_istream& operator>>(basic_istream& (*__pf)(basic_istream&))
{ return __pf(*this); }
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream& operator>>(basic_ios<char_type, traits_type>&
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
(*__pf)(basic_ios<char_type, traits_type>&))
{ __pf(*this); return *this; }
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
basic_istream& operator>>(ios_base& (*__pf)(ios_base&))
{ __pf(*this); return *this; }
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream& operator>>(basic_streambuf<char_type, traits_type>* __sb);
basic_istream& operator>>(bool& __n);
basic_istream& operator>>(short& __n);
basic_istream& operator>>(unsigned short& __n);
basic_istream& operator>>(int& __n);
basic_istream& operator>>(unsigned int& __n);
basic_istream& operator>>(long& __n);
basic_istream& operator>>(unsigned long& __n);
basic_istream& operator>>(long long& __n);
basic_istream& operator>>(unsigned long long& __n);
basic_istream& operator>>(float& __f);
basic_istream& operator>>(double& __f);
basic_istream& operator>>(long double& __f);
basic_istream& operator>>(void*& __p);
// 27.7.1.3 Unformatted input:
visibility-decoration. llvm-svn: 114545
2010-09-22 23:29:08 +08:00
_LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
streamsize gcount() const {return __gc_;}
int_type get();
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
basic_istream& get(char_type& __c) {
int_type __ch = get();
if (__ch != traits_type::eof())
__c = traits_type::to_char_type(__ch);
return *this;
}
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
basic_istream& get(char_type* __s, streamsize __n)
{ return get(__s, __n, this->widen('\n')); }
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream& get(char_type* __s, streamsize __n, char_type __dlm);
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
basic_istream& get(basic_streambuf<char_type, traits_type>& __sb)
{ return get(__sb, this->widen('\n')); }
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream& get(basic_streambuf<char_type, traits_type>& __sb, char_type __dlm);
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
basic_istream& getline(char_type* __s, streamsize __n)
{ return getline(__s, __n, this->widen('\n')); }
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream& getline(char_type* __s, streamsize __n, char_type __dlm);
basic_istream& ignore(streamsize __n = 1, int_type __dlm = traits_type::eof());
int_type peek();
basic_istream& read (char_type* __s, streamsize __n);
streamsize readsome(char_type* __s, streamsize __n);
basic_istream& putback(char_type __c);
basic_istream& unget();
int sync();
pos_type tellg();
basic_istream& seekg(pos_type __pos);
basic_istream& seekg(off_type __off, ios_base::seekdir __dir);
};
template <class _CharT, class _Traits>
[NFC] Rename _LIBCPP_TYPE_VIS_ONLY to _LIBCPP_TEMPLATE_VIS The name _LIBCPP_TYPE_VIS_ONLY is no longer accurate because both _LIBCPP_TYPE_VIS and _LIBCPP_TYPE_VIS_ONLY expand to __attribute__((__type_visibility__)) with Clang. The only remaining difference is that _LIBCPP_TYPE_VIS_ONLY can be applied to templates whereas _LIBCPP_TYPE_VIS cannot (due to dllimport/dllexport not being allowed on templates). This patch renames _LIBCPP_TYPE_VIS_ONLY to _LIBCPP_TEMPLATE_VIS. llvm-svn: 291035
2017-01-05 07:56:00 +08:00
class _LIBCPP_TEMPLATE_VIS basic_istream<_CharT, _Traits>::sentry
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
bool __ok_;
sentry(const sentry&); // = delete;
sentry& operator=(const sentry&); // = delete;
public:
explicit sentry(basic_istream<_CharT, _Traits>& __is, bool __noskipws = false);
// ~sentry() = default;
visibility-decoration. llvm-svn: 114545
2010-09-22 23:29:08 +08:00
_LIBCPP_INLINE_VISIBILITY
Modernize conversion to bool to the explicit bool conversion operator (library wide). This fixes http://llvm.org/bugs/show_bug.cgi?id=12058. llvm-svn: 151088
2012-02-22 05:46:43 +08:00
_LIBCPP_EXPLICIT
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
operator bool() const {return __ok_;}
};
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>::sentry::sentry(basic_istream<_CharT, _Traits>& __is,
bool __noskipws)
: __ok_(false)
{
if (__is.good())
{
if (__is.tie())
__is.tie()->flush();
if (!__noskipws && (__is.flags() & ios_base::skipws))
{
Further macro protection by replacing _[A-Z] with _[A-Z]p llvm-svn: 145410
2011-11-30 02:15:50 +08:00
typedef istreambuf_iterator<_CharT, _Traits> _Ip;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
const ctype<_CharT>& __ct = use_facet<ctype<_CharT> >(__is.getloc());
Further macro protection by replacing _[A-Z] with _[A-Z]p llvm-svn: 145410
2011-11-30 02:15:50 +08:00
_Ip __i(__is);
_Ip __eof;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
for (; __i != __eof; ++__i)
if (!__ct.is(__ct.space, *__i))
break;
if (__i == __eof)
__is.setstate(ios_base::failbit | ios_base::eofbit);
}
__ok_ = __is.good();
}
else
__is.setstate(ios_base::failbit);
}
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#ifndef _LIBCPP_CXX03_LANG
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>::basic_istream(basic_istream&& __rhs)
: __gc_(__rhs.__gc_)
{
__rhs.__gc_ = 0;
this->move(__rhs);
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator=(basic_istream&& __rhs)
{
swap(__rhs);
return *this;
}
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#endif // _LIBCPP_CXX03_LANG
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>::~basic_istream()
{
}
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
template <class _Tp, class _CharT, class _Traits>
_LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream<_CharT, _Traits>&
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
__input_arithmetic(basic_istream<_CharT, _Traits>& __is, _Tp& __n) {
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
typename basic_istream<_CharT, _Traits>::sentry __s(__is);
if (__s)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
typedef istreambuf_iterator<_CharT, _Traits> _Ip;
typedef num_get<_CharT, _Ip> _Fp;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
use_facet<_Fp>(__is.getloc()).get(_Ip(__is), _Ip(), __is, __state, __n);
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
__is.__setstate_nothrow(__state);
if (__is.exceptions() & ios_base::badbit)
{
throw;
}
}
#endif
__is.setstate(__state);
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
}
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return __is;
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(unsigned short& __n)
{
return _VSTD::__input_arithmetic<unsigned short>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(unsigned int& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<unsigned int>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(long& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<long>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(unsigned long& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<unsigned long>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(long long& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<long long>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(unsigned long long& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<unsigned long long>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(float& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<float>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(double& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<double>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(long double& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<long double>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(bool& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<bool>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(void*& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic<void*>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
template <class _Tp, class _CharT, class _Traits>
_LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream<_CharT, _Traits>&
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
__input_arithmetic_with_numeric_limits(basic_istream<_CharT, _Traits>& __is, _Tp& __n) {
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
typename basic_istream<_CharT, _Traits>::sentry __s(__is);
if (__s)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
typedef istreambuf_iterator<_CharT, _Traits> _Ip;
typedef num_get<_CharT, _Ip> _Fp;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
long __temp;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
use_facet<_Fp>(__is.getloc()).get(_Ip(__is), _Ip(), __is, __state, __temp);
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
if (__temp < numeric_limits<_Tp>::min())
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
__n = numeric_limits<_Tp>::min();
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
else if (__temp > numeric_limits<_Tp>::max())
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
__n = numeric_limits<_Tp>::max();
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
else
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
__n = static_cast<_Tp>(__temp);
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
__is.__setstate_nothrow(__state);
if (__is.exceptions() & ios_base::badbit)
{
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__is.setstate(__state);
}
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return __is;
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(short& __n)
{
return _VSTD::__input_arithmetic_with_numeric_limits<short>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(int& __n)
{
[libc++] Factor duplicate code into function templates Summary: The exact same code was replicated 11 times for implementing the basic_istream input operators (those that don't use numeric_limits). The same code was also duplicated twice for implementing the basic_istream input operators that take numeric_limits into account. This commit factors the common code into function templates to avoid the duplication. Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49808 llvm-svn: 337955
2018-07-26 03:40:01 +08:00
return _VSTD::__input_arithmetic_with_numeric_limits<int>(*this, __n);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template<class _CharT, class _Traits>
[libc++] Implement P0487R1 - Fixing operator>>(basic_istream&, CharT*) Summary: Avoid buffer overflow by replacing the pointer interface with an array reference interface in C++2a. Tentatively ready on Batavia2018. https://wg21.link/lwg2499 https://wg21.link/p0487 Reviewers: mclow.lists, ldionne, EricWF Reviewed By: ldionne Subscribers: libcxx-commits, cfe-commits, christof Differential Revision: https://reviews.llvm.org/D51268 llvm-svn: 347377
2018-11-21 11:30:10 +08:00
_LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream<_CharT, _Traits>&
[libc++] Implement P0487R1 - Fixing operator>>(basic_istream&, CharT*) Summary: Avoid buffer overflow by replacing the pointer interface with an array reference interface in C++2a. Tentatively ready on Batavia2018. https://wg21.link/lwg2499 https://wg21.link/p0487 Reviewers: mclow.lists, ldionne, EricWF Reviewed By: ldionne Subscribers: libcxx-commits, cfe-commits, christof Differential Revision: https://reviews.llvm.org/D51268 llvm-svn: 347377
2018-11-21 11:30:10 +08:00
__input_c_string(basic_istream<_CharT, _Traits>& __is, _CharT* __p, size_t __n)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
typename basic_istream<_CharT, _Traits>::sentry __sen(__is);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif
[libc++] Remove use of auto with actual type For compatibility with C++03. llvm-svn: 357512
2019-04-03 03:20:47 +08:00
_CharT* __s = __p;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
const ctype<_CharT>& __ct = use_facet<ctype<_CharT> >(__is.getloc());
[libc++] Implement P0487R1 - Fixing operator>>(basic_istream&, CharT*) Summary: Avoid buffer overflow by replacing the pointer interface with an array reference interface in C++2a. Tentatively ready on Batavia2018. https://wg21.link/lwg2499 https://wg21.link/p0487 Reviewers: mclow.lists, ldionne, EricWF Reviewed By: ldionne Subscribers: libcxx-commits, cfe-commits, christof Differential Revision: https://reviews.llvm.org/D51268 llvm-svn: 347377
2018-11-21 11:30:10 +08:00
while (__s != __p + (__n-1))
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
typename _Traits::int_type __i = __is.rdbuf()->sgetc();
if (_Traits::eq_int_type(__i, _Traits::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
_CharT __ch = _Traits::to_char_type(__i);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
if (__ct.is(__ct.space, __ch))
break;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
*__s++ = __ch;
__is.rdbuf()->sbumpc();
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
*__s = _CharT();
__is.width(0);
[libc++] Implement P0487R1 - Fixing operator>>(basic_istream&, CharT*) Summary: Avoid buffer overflow by replacing the pointer interface with an array reference interface in C++2a. Tentatively ready on Batavia2018. https://wg21.link/lwg2499 https://wg21.link/p0487 Reviewers: mclow.lists, ldionne, EricWF Reviewed By: ldionne Subscribers: libcxx-commits, cfe-commits, christof Differential Revision: https://reviews.llvm.org/D51268 llvm-svn: 347377
2018-11-21 11:30:10 +08:00
if (__s == __p)
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
__is.__setstate_nothrow(__state);
if (__is.exceptions() & ios_base::badbit)
{
throw;
}
}
#endif
__is.setstate(__state);
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return __is;
}
[libc++] Implement P0487R1 - Fixing operator>>(basic_istream&, CharT*) Summary: Avoid buffer overflow by replacing the pointer interface with an array reference interface in C++2a. Tentatively ready on Batavia2018. https://wg21.link/lwg2499 https://wg21.link/p0487 Reviewers: mclow.lists, ldionne, EricWF Reviewed By: ldionne Subscribers: libcxx-commits, cfe-commits, christof Differential Revision: https://reviews.llvm.org/D51268 llvm-svn: 347377
2018-11-21 11:30:10 +08:00
#if _LIBCPP_STD_VER > 17
template<class _CharT, class _Traits, size_t _Np>
inline _LIBCPP_INLINE_VISIBILITY
basic_istream<_CharT, _Traits>&
operator>>(basic_istream<_CharT, _Traits>& __is, _CharT (&__buf)[_Np])
{
[libc++] Remove use of auto with actual type For compatibility with C++03. llvm-svn: 357512
2019-04-03 03:20:47 +08:00
size_t __n = _Np;
[libc++] Implement P0487R1 - Fixing operator>>(basic_istream&, CharT*) Summary: Avoid buffer overflow by replacing the pointer interface with an array reference interface in C++2a. Tentatively ready on Batavia2018. https://wg21.link/lwg2499 https://wg21.link/p0487 Reviewers: mclow.lists, ldionne, EricWF Reviewed By: ldionne Subscribers: libcxx-commits, cfe-commits, christof Differential Revision: https://reviews.llvm.org/D51268 llvm-svn: 347377
2018-11-21 11:30:10 +08:00
if (__is.width() > 0)
__n = _VSTD::min(size_t(__is.width()), _Np);
return _VSTD::__input_c_string(__is, __buf, __n);
}
template<class _Traits, size_t _Np>
inline _LIBCPP_INLINE_VISIBILITY
basic_istream<char, _Traits>&
operator>>(basic_istream<char, _Traits>& __is, unsigned char (&__buf)[_Np])
{
return __is >> (char(&)[_Np])__buf;
}
template<class _Traits, size_t _Np>
inline _LIBCPP_INLINE_VISIBILITY
basic_istream<char, _Traits>&
operator>>(basic_istream<char, _Traits>& __is, signed char (&__buf)[_Np])
{
return __is >> (char(&)[_Np])__buf;
}
#else
template<class _CharT, class _Traits>
inline _LIBCPP_INLINE_VISIBILITY
basic_istream<_CharT, _Traits>&
operator>>(basic_istream<_CharT, _Traits>& __is, _CharT* __s)
{
streamsize __n = __is.width();
if (__n <= 0)
__n = numeric_limits<streamsize>::max() / sizeof(_CharT) - 1;
return _VSTD::__input_c_string(__is, __s, size_t(__n));
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template<class _Traits>
Revert "Remove visibility attributes from out-of-class method definitions in iostreams." llvm-svn: 257193
2016-01-09 03:21:02 +08:00
inline _LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream<char, _Traits>&
operator>>(basic_istream<char, _Traits>& __is, unsigned char* __s)
{
return __is >> (char*)__s;
}
template<class _Traits>
Revert "Remove visibility attributes from out-of-class method definitions in iostreams." llvm-svn: 257193
2016-01-09 03:21:02 +08:00
inline _LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream<char, _Traits>&
operator>>(basic_istream<char, _Traits>& __is, signed char* __s)
{
return __is >> (char*)__s;
}
[libc++] Implement P0487R1 - Fixing operator>>(basic_istream&, CharT*) Summary: Avoid buffer overflow by replacing the pointer interface with an array reference interface in C++2a. Tentatively ready on Batavia2018. https://wg21.link/lwg2499 https://wg21.link/p0487 Reviewers: mclow.lists, ldionne, EricWF Reviewed By: ldionne Subscribers: libcxx-commits, cfe-commits, christof Differential Revision: https://reviews.llvm.org/D51268 llvm-svn: 347377
2018-11-21 11:30:10 +08:00
#endif // _LIBCPP_STD_VER > 17
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
operator>>(basic_istream<_CharT, _Traits>& __is, _CharT& __c)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
typename basic_istream<_CharT, _Traits>::sentry __sen(__is);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif
Fixed two problems found by Chris Jefferson: Made operator>> for char consistent with gcc. Opened an LWG issue on this one. 2) Renamed some private typedefs which are causing boost grief. llvm-svn: 126576
2011-02-28 02:02:02 +08:00
typename _Traits::int_type __i = __is.rdbuf()->sbumpc();
if (_Traits::eq_int_type(__i, _Traits::eof()))
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit | ios_base::failbit;
Fixed two problems found by Chris Jefferson: Made operator>> for char consistent with gcc. Opened an LWG issue on this one. 2) Renamed some private typedefs which are causing boost grief. llvm-svn: 126576
2011-02-28 02:02:02 +08:00
else
__c = _Traits::to_char_type(__i);
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
__is.__setstate_nothrow(__state);
if (__is.exceptions() & ios_base::badbit)
{
throw;
}
}
#endif
__is.setstate(__state);
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return __is;
}
template<class _Traits>
Revert "Remove visibility attributes from out-of-class method definitions in iostreams." llvm-svn: 257193
2016-01-09 03:21:02 +08:00
inline _LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream<char, _Traits>&
operator>>(basic_istream<char, _Traits>& __is, unsigned char& __c)
{
return __is >> (char&)__c;
}
template<class _Traits>
Revert "Remove visibility attributes from out-of-class method definitions in iostreams." llvm-svn: 257193
2016-01-09 03:21:02 +08:00
inline _LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_istream<char, _Traits>&
operator>>(basic_istream<char, _Traits>& __is, signed char& __c)
{
return __is >> (char&)__c;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::operator>>(basic_streambuf<char_type, traits_type>* __sb)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __s(*this, true);
if (__s)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
if (__sb)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
try
{
#endif // _LIBCPP_NO_EXCEPTIONS
while (true)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
typename traits_type::int_type __i = this->rdbuf()->sgetc();
if (traits_type::eq_int_type(__i, _Traits::eof()))
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
break;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
if (traits_type::eq_int_type(
__sb->sputc(traits_type::to_char_type(__i)),
traits_type::eof()))
break;
++__gc_;
this->rdbuf()->sbumpc();
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
if (__gc_ == 0)
__state |= ios_base::failbit;
#ifndef _LIBCPP_NO_EXCEPTIONS
}
catch (...)
{
__state |= ios_base::badbit;
if (__gc_ == 0)
__state |= ios_base::failbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::failbit || this->exceptions() & ios_base::badbit)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
throw;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
else
{
__state |= ios_base::failbit;
}
this->setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
return *this;
}
template<class _CharT, class _Traits>
typename basic_istream<_CharT, _Traits>::int_type
basic_istream<_CharT, _Traits>::get()
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
int_type __r = traits_type::eof();
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __s(*this, true);
if (__s)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
__r = this->rdbuf()->sbumpc();
if (traits_type::eq_int_type(__r, traits_type::eof()))
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit | ios_base::eofbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
else
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
__gc_ = 1;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
this->__setstate_nothrow(this->rdstate() | ios_base::badbit);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
#endif
this->setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
return __r;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::get(char_type* __s, streamsize __n, char_type __dlm)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
if (__n > 0)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
while (__gc_ < __n-1)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
int_type __i = this->rdbuf()->sgetc();
if (traits_type::eq_int_type(__i, traits_type::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
char_type __ch = traits_type::to_char_type(__i);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
if (traits_type::eq(__ch, __dlm))
break;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
*__s++ = __ch;
++__gc_;
this->rdbuf()->sbumpc();
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
if (__gc_ == 0)
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. I will submit a paper in San Diego to clarify the Standard such that the interpretation used in this commit (and other implementations) is the only possible one. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357531
2019-04-03 05:43:07 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
if (__n > 0)
*__s = char_type();
throw;
}
}
#endif
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. I will submit a paper in San Diego to clarify the Standard such that the interpretation used in this commit (and other implementations) is the only possible one. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357531
2019-04-03 05:43:07 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
else
{
__state |= ios_base::failbit;
}
[libcxx] Make std::basic_istream::get 0-terminate input array in case of error. It covers the cases when the sentry object returns false and when an exception was thrown. Corresponding standard paragraph is C++14 [istream.unformatted]p9: [...] In any case, if n is greater than zero it then stores a null character into the next successive location of the array. rdar://problem/35566567 Reviewers: EricWF, mclow.lists Reviewed By: mclow.lists Subscribers: cfe-commits Differential Revision: https://reviews.llvm.org/D40677 llvm-svn: 322326
2018-01-12 07:23:49 +08:00
if (__n > 0)
*__s = char_type();
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
if (__n > 0)
*__s = char_type();
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::get(basic_streambuf<char_type, traits_type>& __sb,
char_type __dlm)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
try
{
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
while (true)
{
typename traits_type::int_type __i = this->rdbuf()->sgetc();
if (traits_type::eq_int_type(__i, traits_type::eof()))
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
break;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
char_type __ch = traits_type::to_char_type(__i);
if (traits_type::eq(__ch, __dlm))
break;
if (traits_type::eq_int_type(__sb.sputc(__ch), traits_type::eof()))
break;
++__gc_;
this->rdbuf()->sbumpc();
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
}
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
// according to the spec, exceptions here are caught but not rethrown
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
if (__gc_ == 0)
__state |= ios_base::failbit;
this->setstate(__state);
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::getline(char_type* __s, streamsize __n, char_type __dlm)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
while (true)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
typename traits_type::int_type __i = this->rdbuf()->sgetc();
if (traits_type::eq_int_type(__i, traits_type::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
char_type __ch = traits_type::to_char_type(__i);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
if (traits_type::eq(__ch, __dlm))
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
{
this->rdbuf()->sbumpc();
++__gc_;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
break;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
}
if (__gc_ >= __n-1)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
break;
}
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
*__s++ = __ch;
this->rdbuf()->sbumpc();
++__gc_;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
if (__n > 0)
*__s = char_type();
if (__gc_ == 0)
__state |= ios_base::failbit;
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
if (__n > 0)
*__s = char_type();
if (__gc_ == 0)
__state |= ios_base::failbit;
this->setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::ignore(streamsize __n, int_type __dlm)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
if (__n == numeric_limits<streamsize>::max())
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
while (true)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
typename traits_type::int_type __i = this->rdbuf()->sbumpc();
if (traits_type::eq_int_type(__i, traits_type::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
++__gc_;
In istream::ignore, check the delimeter as an int_type, not as a char_type, so as to correctly handle EOF. This fixes http://llvm.org/bugs/show_bug.cgi?id=16427 llvm-svn: 185298
2013-07-01 08:37:50 +08:00
if (traits_type::eq_int_type(__i, __dlm))
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
break;
}
}
else
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
while (__gc_ < __n)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
typename traits_type::int_type __i = this->rdbuf()->sbumpc();
if (traits_type::eq_int_type(__i, traits_type::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
++__gc_;
In istream::ignore, check the delimeter as an int_type, not as a char_type, so as to correctly handle EOF. This fixes http://llvm.org/bugs/show_bug.cgi?id=16427 llvm-svn: 185298
2013-07-01 08:37:50 +08:00
if (traits_type::eq_int_type(__i, __dlm))
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
break;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template<class _CharT, class _Traits>
typename basic_istream<_CharT, _Traits>::int_type
basic_istream<_CharT, _Traits>::peek()
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
int_type __r = traits_type::eof();
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__r = this->rdbuf()->sgetc();
peek should set eofbit if sgetc() returns eof. llvm-svn: 167238
2012-11-02 01:32:07 +08:00
if (traits_type::eq_int_type(__r, traits_type::eof()))
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return __r;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::read(char_type* __s, streamsize __n)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
Have basic_istream::read call sgetn intead of sbumpc individual characters. This addresses http://llvm.org/bugs/show_bug.cgi?id=15427. llvm-svn: 176573
2013-03-07 03:27:56 +08:00
__gc_ = this->rdbuf()->sgetn(__s, __n);
if (__gc_ != __n)
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit | ios_base::eofbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
#endif // _LIBCPP_NO_EXCEPTIONS
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
else
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template<class _CharT, class _Traits>
streamsize
basic_istream<_CharT, _Traits>::readsome(char_type* __s, streamsize __n)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
Hyeon-Bin Jeong: readsome() need to reset gcount to zero. This fixes http://llvm.org/bugs/show_bug.cgi?id=14670. llvm-svn: 170703
2012-12-20 23:40:28 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
Constuct a sentry object in istream::readsome, and handle failures appropriately. Fixes PR#28217. llvm-svn: 275280
2016-07-14 00:58:48 +08:00
streamsize __c = this->rdbuf()->in_avail();
switch (__c)
{
case -1:
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Constuct a sentry object in istream::readsome, and handle failures appropriately. Fixes PR#28217. llvm-svn: 275280
2016-07-14 00:58:48 +08:00
break;
case 0:
break;
default:
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__n = _VSTD::min(__c, __n);
__gc_ = this->rdbuf()->sgetn(__s, __n);
if (__gc_ != __n)
__state |= ios_base::failbit | ios_base::eofbit;
Constuct a sentry object in istream::readsome, and handle failures appropriately. Fixes PR#28217. llvm-svn: 275280
2016-07-14 00:58:48 +08:00
break;
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
#endif // _LIBCPP_NO_EXCEPTIONS
Constuct a sentry object in istream::readsome, and handle failures appropriately. Fixes PR#28217. llvm-svn: 275280
2016-07-14 00:58:48 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
else
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
return __gc_;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::putback(char_type __c)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = this->rdstate() & ~ios_base::eofbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->clear(__state);
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. I will submit a paper in San Diego to clarify the Standard such that the interpretation used in this commit (and other implementations) is the only possible one. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357531
2019-04-03 05:43:07 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Replace several uses of 0 by nullptr Differential Revision: https://reviews.llvm.org/D43159
2020-11-25 01:53:53 +08:00
if (this->rdbuf() == nullptr || this->rdbuf()->sputbackc(__c) == traits_type::eof())
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::badbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
#endif // _LIBCPP_NO_EXCEPTIONS
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
else
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::unget()
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = this->rdstate() & ~ios_base::eofbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__gc_ = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->clear(__state);
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. I will submit a paper in San Diego to clarify the Standard such that the interpretation used in this commit (and other implementations) is the only possible one. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357531
2019-04-03 05:43:07 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Replace several uses of 0 by nullptr Differential Revision: https://reviews.llvm.org/D43159
2020-11-25 01:53:53 +08:00
if (this->rdbuf() == nullptr || this->rdbuf()->sungetc() == traits_type::eof())
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::badbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
#endif // _LIBCPP_NO_EXCEPTIONS
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
else
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template<class _CharT, class _Traits>
int
basic_istream<_CharT, _Traits>::sync()
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
int __r = 0;
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Replace several uses of 0 by nullptr Differential Revision: https://reviews.llvm.org/D43159
2020-11-25 01:53:53 +08:00
if (this->rdbuf() == nullptr)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return -1;
if (this->rdbuf()->pubsync() == -1)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::badbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return -1;
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return __r;
}
template<class _CharT, class _Traits>
typename basic_istream<_CharT, _Traits>::pos_type
basic_istream<_CharT, _Traits>::tellg()
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
pos_type __r(-1);
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
{
Fixing whitespace problems llvm-svn: 111750
2010-08-22 08:02:43 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__r = this->rdbuf()->pubseekoff(0, ios_base::cur, ios_base::in);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return __r;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::seekg(pos_type __pos)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = this->rdstate() & ~ios_base::eofbit;
this->clear(__state);
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. I will submit a paper in San Diego to clarify the Standard such that the interpretation used in this commit (and other implementations) is the only possible one. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357531
2019-04-03 05:43:07 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
if (this->rdbuf()->pubseekpos(__pos, ios_base::in) == pos_type(-1))
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template<class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
basic_istream<_CharT, _Traits>::seekg(off_type __off, ios_base::seekdir __dir)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = this->rdstate() & ~ios_base::eofbit;
this->clear(__state);
sentry __sen(*this, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. I will submit a paper in San Diego to clarify the Standard such that the interpretation used in this commit (and other implementations) is the only possible one. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357531
2019-04-03 05:43:07 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
if (this->rdbuf()->pubseekoff(__off, __dir, ios_base::in) == pos_type(-1))
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
this->__setstate_nothrow(__state);
if (this->exceptions() & ios_base::badbit)
{
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
this->setstate(__state);
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return *this;
}
template <class _CharT, class _Traits>
basic_istream<_CharT, _Traits>&
ws(basic_istream<_CharT, _Traits>& __is)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
typename basic_istream<_CharT, _Traits>::sentry __sen(__is, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
const ctype<_CharT>& __ct = use_facet<ctype<_CharT> >(__is.getloc());
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
while (true)
{
typename _Traits::int_type __i = __is.rdbuf()->sgetc();
if (_Traits::eq_int_type(__i, _Traits::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
if (!__ct.is(__ct.space, _Traits::to_char_type(__i)))
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
break;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
__is.rdbuf()->sbumpc();
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
__is.__setstate_nothrow(__state);
if (__is.exceptions() & ios_base::badbit)
{
throw;
}
}
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#endif // _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__is.setstate(__state);
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return __is;
}
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#ifndef _LIBCPP_CXX03_LANG
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
[libc++] Re-apply fdc41e11f (LWG1203) without breaking the C++11 build fdc41e11f was reverted in e46c1def5 because it broke the C++11 build. We shouldn't be using enable_if_t in C++11, instead we must use enable_if<...>::type.
2020-09-23 20:49:00 +08:00
template <class _Stream, class _Tp, class = void>
struct __is_istreamable : false_type { };
template <class _Stream, class _Tp>
struct __is_istreamable<_Stream, _Tp, decltype(
_VSTD::declval<_Stream>() >> _VSTD::declval<_Tp>(), void()
)> : true_type { };
template <class _Stream, class _Tp, class = typename enable_if<
_And<is_base_of<ios_base, _Stream>,
__is_istreamable<_Stream&, _Tp&&>>::value
>::type>
_LIBCPP_INLINE_VISIBILITY
_Stream&& operator>>(_Stream&& __is, _Tp&& __x)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Implement LWG2328. Rvalue stream extraction should perfect forward. llvm-svn: 276545
2016-07-24 12:07:22 +08:00
__is >> _VSTD::forward<_Tp>(__x);
[libc++] Re-apply fdc41e11f (LWG1203) without breaking the C++11 build fdc41e11f was reverted in e46c1def5 because it broke the C++11 build. We shouldn't be using enable_if_t in C++11, instead we must use enable_if<...>::type.
2020-09-23 20:49:00 +08:00
return _VSTD::move(__is);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#endif // _LIBCPP_CXX03_LANG
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template <class _CharT, class _Traits>
[NFC] Rename _LIBCPP_TYPE_VIS_ONLY to _LIBCPP_TEMPLATE_VIS The name _LIBCPP_TYPE_VIS_ONLY is no longer accurate because both _LIBCPP_TYPE_VIS and _LIBCPP_TYPE_VIS_ONLY expand to __attribute__((__type_visibility__)) with Clang. The only remaining difference is that _LIBCPP_TYPE_VIS_ONLY can be applied to templates whereas _LIBCPP_TYPE_VIS cannot (due to dllimport/dllexport not being allowed on templates). This patch renames _LIBCPP_TYPE_VIS_ONLY to _LIBCPP_TEMPLATE_VIS. llvm-svn: 291035
2017-01-05 07:56:00 +08:00
class _LIBCPP_TEMPLATE_VIS basic_iostream
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
: public basic_istream<_CharT, _Traits>,
public basic_ostream<_CharT, _Traits>
{
public:
// types:
typedef _CharT char_type;
typedef _Traits traits_type;
typedef typename traits_type::int_type int_type;
typedef typename traits_type::pos_type pos_type;
typedef typename traits_type::off_type off_type;
// constructor/destructor
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
explicit basic_iostream(basic_streambuf<char_type, traits_type>* __sb)
: basic_istream<_CharT, _Traits>(__sb)
{}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
virtual ~basic_iostream();
protected:
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#ifndef _LIBCPP_CXX03_LANG
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
inline _LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_iostream(basic_iostream&& __rhs);
// assign/swap
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
inline _LIBCPP_INLINE_VISIBILITY
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_iostream& operator=(basic_iostream&& __rhs);
#endif
[libcxx] Remove _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY That macro has been defined to _LIBCPP_HIDE_FROM_ABI_AFTER_V1 for many weeks now, so we're actually replacing uses of it for uses of _LIBCPP_HIDE_FROM_ABI_AFTER_V1 directly. This should not change or break anything since the two macros are 100% equivalent, unless somebody is (incorrectly!) relying on _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY being defined. llvm-svn: 344641
2018-10-17 03:26:23 +08:00
inline _LIBCPP_HIDE_FROM_ABI_AFTER_V1
[libc++] Add _LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY to support GCC ABI compatibility Summary: GCC and Clang handle visibility attributes on the out-of-line definition of externally instantiated templates differently. For example in the reproducer below Clang will emit both 'foo' and 'bar' with default visibility while GCC only emits a non-hidden 'foo'. ``` // RUN: g++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out // RUN: clang++ -std=c++11 -shared -O3 test.cpp && sym_extract.py a.out #define INLINE_VISIBILITY __attribute__((visibility("hidden"), always_inline)) template <class T> struct Foo { void foo(); void bar(); }; template <class T> void Foo<T>::foo() {} template <class T> inline INLINE_VISIBILITY void Foo<T>::bar() {} template struct Foo<int>; ``` This difference creates ABI incompatibilities between Clang and GCC built dylibs. Specifically GCC built dylibs lack definitions for various member functions of `basic_string`, `basic_istream`, `basic_ostream`, `basic_iostream`, and `basic_streambuf` (All of these types are externally instantiated). Surprisingly these missing symbols don't cause many problems because the functions are marked `always_inline` therefore the dylib definition is rarely needed. However when an out-of-line definition is required then GCC built dylibs will fail to link. For example [GCC built dylibs cannot build Clang](http://stackoverflow.com/questions/39454262/clang-build-errors). This patch works around this issue by adding `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` which is used to mark externally instantiated member functions as always inline. When building the library `_LIBCPP_EXTERN_TEMPLATE_INLINE_VISIBILITY` sets the symbol's visibility to "default" instead of "hidden", otherwise it acts exactly the same as `_LIBCPP_INLINE_VISIBILITY`. After applying this patch GCC dylibs now contain: * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7sungetcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5gbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7sungetcEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9sputbackcEc` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setpEPcS4_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6snextcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcm` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE9pubsetbufEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_9basic_iosIwS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5pbumpEi` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE7getlineEPcl` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetcEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8in_availEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE6sbumpcEv` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getERc` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6snextcEv` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEmw` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE7getlineEPwl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5tellpEv` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getERw` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEmc` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE7pubsyncEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEE3getEPcl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRNS_9basic_iosIcS2_EES6_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE7pubsyncEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputcEc` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpExNS_8ios_base7seekdirE` * `_ZNKSt3__115basic_streambufIcNS_11char_traitsIcEEE6getlocEv` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5gbumpEi` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5seekpENS_4fposI11__mbstate_tEE` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE5tellpEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRS3_S4_E` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEE3getEPwl` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEElsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE4setgEPcS4_S4_` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setgEPwS4_S4_` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEEC1EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE8pubimbueERKNS_6localeE` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE4swapERS3_` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEEC2EPNS_15basic_streambufIwS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5pbumpEi` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetcEv` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEE4swapERS3_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE10pubseekposENS_4fposI11__mbstate_tEEj` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sputnEPKcl` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEE5seekpExNS_8ios_base7seekdirE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sgetnEPwl` * `_ZNSt3__113basic_ostreamIwNS_11char_traitsIwEEElsEPFRNS_8ios_baseES5_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE4setpEPwS4_` * `_ZNSt3__115basic_streambufIcNS_11char_traitsIcEEE5sgetnEPcl` * `_ZNKSt3__115basic_streambufIwNS_11char_traitsIwEEE6getlocEv` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8pubimbueERKNS_6localeE` * `_ZNSt3__114basic_iostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE8in_availEv` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6__initEPKcmm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE6sbumpcEv` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE10pubseekoffExNS_8ios_base7seekdirEj` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC2EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__113basic_istreamIwNS_11char_traitsIwEEErsEPFRS3_S4_E` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9sputbackcEw` * `_ZNSt3__112basic_stringIwNS_11char_traitsIwEENS_9allocatorIwEEE6__initEPKwm` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputnEPKwl` * `_ZNSt3__113basic_istreamIcNS_11char_traitsIcEEErsEPFRS3_S4_E` * `_ZNSt3__113basic_ostreamIcNS_11char_traitsIcEEEC1EPNS_15basic_streambufIcS2_EE` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE9pubsetbufEPwl` * `_ZNSt3__115basic_streambufIwNS_11char_traitsIwEEE5sputcEw` This patch has no effect on Clang based builds. Reviewers: mclow.lists, eugenis, danalbert, jroelofs, EricWF Subscribers: beanz, cfe-commits, mgorny Differential Revision: https://reviews.llvm.org/D24600 llvm-svn: 281681
2016-09-16 08:00:48 +08:00
void swap(basic_iostream& __rhs)
{ basic_istream<char_type, traits_type>::swap(__rhs); }
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
};
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#ifndef _LIBCPP_CXX03_LANG
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template <class _CharT, class _Traits>
basic_iostream<_CharT, _Traits>::basic_iostream(basic_iostream&& __rhs)
_STD -> _VSTD to avoid macro clash on windows llvm-svn: 134190
2011-07-01 05:18:19 +08:00
: basic_istream<_CharT, _Traits>(_VSTD::move(__rhs))
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
}
template <class _CharT, class _Traits>
basic_iostream<_CharT, _Traits>&
basic_iostream<_CharT, _Traits>::operator=(basic_iostream&& __rhs)
{
swap(__rhs);
return *this;
}
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#endif // _LIBCPP_CXX03_LANG
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template <class _CharT, class _Traits>
basic_iostream<_CharT, _Traits>::~basic_iostream()
{
}
template<class _CharT, class _Traits, class _Allocator>
basic_istream<_CharT, _Traits>&
operator>>(basic_istream<_CharT, _Traits>& __is,
basic_string<_CharT, _Traits, _Allocator>& __str)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
typename basic_istream<_CharT, _Traits>::sentry __sen(__is);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__str.clear();
streamsize __n = __is.width();
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
if (__n <= 0)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__n = __str.max_size();
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
if (__n <= 0)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__n = numeric_limits<streamsize>::max();
streamsize __c = 0;
const ctype<_CharT>& __ct = use_facet<ctype<_CharT> >(__is.getloc());
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
while (__c < __n)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
typename _Traits::int_type __i = __is.rdbuf()->sgetc();
if (_Traits::eq_int_type(__i, _Traits::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
_CharT __ch = _Traits::to_char_type(__i);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
if (__ct.is(__ct.space, __ch))
break;
__str.push_back(__ch);
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
++__c;
__is.rdbuf()->sbumpc();
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
__is.width(0);
if (__c == 0)
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
__is.__setstate_nothrow(__state);
if (__is.exceptions() & ios_base::badbit)
{
throw;
}
}
#endif
__is.setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
return __is;
}
template<class _CharT, class _Traits, class _Allocator>
Fixing whitespace problems llvm-svn: 111750
2010-08-22 08:02:43 +08:00
basic_istream<_CharT, _Traits>&
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
getline(basic_istream<_CharT, _Traits>& __is,
basic_string<_CharT, _Traits, _Allocator>& __str, _CharT __dlm)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
typename basic_istream<_CharT, _Traits>::sentry __sen(__is, true);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
__str.clear();
Fix <rdar://problem/10256836> getline of an empty string mistakenly causes failure llvm-svn: 141506
2011-10-09 23:20:46 +08:00
streamsize __extr = 0;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
while (true)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
typename _Traits::int_type __i = __is.rdbuf()->sbumpc();
if (_Traits::eq_int_type(__i, _Traits::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
Fix <rdar://problem/10256836> getline of an empty string mistakenly causes failure llvm-svn: 141506
2011-10-09 23:20:46 +08:00
++__extr;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
_CharT __ch = _Traits::to_char_type(__i);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
if (_Traits::eq(__ch, __dlm))
break;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
__str.push_back(__ch);
if (__str.size() == __str.max_size())
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
break;
}
}
Fix <rdar://problem/10256836> getline of an empty string mistakenly causes failure llvm-svn: 141506
2011-10-09 23:20:46 +08:00
if (__extr == 0)
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
__is.__setstate_nothrow(__state);
if (__is.exceptions() & ios_base::badbit)
{
throw;
}
}
#endif
__is.setstate(__state);
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
}
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
return __is;
}
template<class _CharT, class _Traits, class _Allocator>
Revert "Remove visibility attributes from out-of-class method definitions in iostreams." llvm-svn: 257193
2016-01-09 03:21:02 +08:00
inline _LIBCPP_INLINE_VISIBILITY
Fixing whitespace problems llvm-svn: 111750
2010-08-22 08:02:43 +08:00
basic_istream<_CharT, _Traits>&
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
getline(basic_istream<_CharT, _Traits>& __is,
basic_string<_CharT, _Traits, _Allocator>& __str)
{
return getline(__is, __str, __is.widen('\n'));
}
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#ifndef _LIBCPP_CXX03_LANG
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template<class _CharT, class _Traits, class _Allocator>
Revert "Remove visibility attributes from out-of-class method definitions in iostreams." llvm-svn: 257193
2016-01-09 03:21:02 +08:00
inline _LIBCPP_INLINE_VISIBILITY
Fixing whitespace problems llvm-svn: 111750
2010-08-22 08:02:43 +08:00
basic_istream<_CharT, _Traits>&
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
getline(basic_istream<_CharT, _Traits>&& __is,
basic_string<_CharT, _Traits, _Allocator>& __str, _CharT __dlm)
{
return getline(__is, __str, __dlm);
}
template<class _CharT, class _Traits, class _Allocator>
Revert "Remove visibility attributes from out-of-class method definitions in iostreams." llvm-svn: 257193
2016-01-09 03:21:02 +08:00
inline _LIBCPP_INLINE_VISIBILITY
Fixing whitespace problems llvm-svn: 111750
2010-08-22 08:02:43 +08:00
basic_istream<_CharT, _Traits>&
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
getline(basic_istream<_CharT, _Traits>&& __is,
basic_string<_CharT, _Traits, _Allocator>& __str)
{
return getline(__is, __str, __is.widen('\n'));
}
Cleanup _LIBCPP_HAS_NO_<c++11-feature> in the input.output library llvm-svn: 300626
2017-04-19 07:38:41 +08:00
#endif // _LIBCPP_CXX03_LANG
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
template <class _CharT, class _Traits, size_t _Size>
basic_istream<_CharT, _Traits>&
operator>>(basic_istream<_CharT, _Traits>& __is, bitset<_Size>& __x)
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
ios_base::iostate __state = ios_base::goodbit;
typename basic_istream<_CharT, _Traits>::sentry __sen(__is);
if (__sen)
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
try
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
#endif
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
basic_string<_CharT, _Traits> __str;
const ctype<_CharT>& __ct = use_facet<ctype<_CharT> >(__is.getloc());
fix sign comparison warnings llvm-svn: 290469
2016-12-24 08:24:44 +08:00
size_t __c = 0;
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
_CharT __zero = __ct.widen('0');
_CharT __one = __ct.widen('1');
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
while (__c < _Size)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
{
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
typename _Traits::int_type __i = __is.rdbuf()->sgetc();
if (_Traits::eq_int_type(__i, _Traits::eof()))
{
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::eofbit;
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
break;
}
_CharT __ch = _Traits::to_char_type(__i);
if (!_Traits::eq(__ch, __zero) && !_Traits::eq(__ch, __one))
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
break;
__str.push_back(__ch);
Reimplemented much of <istream> such that single character extractions do not check to see if this is the last character in the stream and thus never set eofbit. This fixes http://llvm.org/bugs/show_bug.cgi?id=10817 . This fix requires a recompiled libc++.dylib to be fully implemented. The recompiled libc++.dylib is ABI compatible with that shipped on Lion. llvm-svn: 138961
2011-09-02 05:02:45 +08:00
++__c;
__is.rdbuf()->sbumpc();
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
__x = bitset<_Size>(__str);
[libc++] Implement LWG 3199 Summary: The resolution of LWG 3199 makes sure that input-streaming into an empty bitset does not set the failbit on the input stream. Reviewers: mclow.lists, EricWF Subscribers: christof, jkorous, dexonsmith, libcxx-commits Tags: #libc Differential Revision: https://reviews.llvm.org/D65105 llvm-svn: 369422
2019-08-21 02:21:06 +08:00
if (_Size > 0 && __c == 0)
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
__state |= ios_base::failbit;
Revert "[libc++] Fix error flags and exceptions propagated from input stream operations" This reverts commits r357533 and r357531, which broke the LLDB data formatters. I'll hold off until we know how to fix the data formatters accordingly. llvm-svn: 357536
2019-04-03 06:21:27 +08:00
#ifndef _LIBCPP_NO_EXCEPTIONS
[libc++] Fix error flags and exceptions propagated from input stream operations Summary: This is a re-application of r357533 and r357531. They had been reverted because we thought the commits broke the LLDB data formatters, but it turns out this was because only r357531 had been included in the CI run. Before this patch, we would only ever throw an exception if the badbit was set on the stream. The Standard is currently very unclear on how exceptions should be propagated and what error flags should be set by the input stream operations. This commit changes libc++ to behave under a different (but valid) interpretation of the Standard. This interpretation of the Standard matches what other implementations are doing. This effectively implements the wording in p1264r0. It hasn't been voted into the Standard yet, however there is wide agreement that the fix is correct and it's just a matter of time before the fix is standardized. PR21586 PR15949 rdar://problem/15347558 Reviewers: mclow.lists, EricWF Subscribers: christof, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D49863 llvm-svn: 357775
2019-04-06 00:33:37 +08:00
}
catch (...)
{
__state |= ios_base::badbit;
__is.__setstate_nothrow(__state);
if (__is.exceptions() & ios_base::badbit)
{
throw;
}
}
#endif
__is.setstate(__state);
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
}
return __is;
}
[libc++] Fix and document visibility attributes for Clang, GCC and Windows. Summary: This patch fixes a number of problems with the visibility macros across GCC (on Unix) and Windows (DLL import/export semantics). All of the visibility macros are now documented under `DesignDocs/VisibilityMacros.rst`. Now I'll no longer forget the subtleties of each! This patch adds two new visibility macros: * `_LIBCPP_ENUM_VIS` for controlling the typeinfo of enum types. Only Clang supports this. * `_LIBCPP_EXTERN_TEMPLATE_TYPE_VIS` for redefining visibility on explicit instantiation declarations. Clang and Windows require this. After applying this patch GCC only emits one -Wattribute warning opposed to 30+. Reviewers: mclow.lists, EricWF Subscribers: beanz, mgorny, cfe-commits Differential Revision: https://reviews.llvm.org/D24602 llvm-svn: 281673
2016-09-16 06:27:07 +08:00
_LIBCPP_EXTERN_TEMPLATE(class _LIBCPP_EXTERN_TEMPLATE_TYPE_VIS basic_istream<char>)
_LIBCPP_EXTERN_TEMPLATE(class _LIBCPP_EXTERN_TEMPLATE_TYPE_VIS basic_istream<wchar_t>)
_LIBCPP_EXTERN_TEMPLATE(class _LIBCPP_EXTERN_TEMPLATE_TYPE_VIS basic_iostream<char>)
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
_LIBCPP_END_NAMESPACE_STD
[Libc++] Use #pragma push_macro/pop_macro to better handle min/max on Windows Summary: This patch improves how libc++ handles min/max macros within the headers. Previously libc++ would undef them and emit a warning. This patch changes libc++ to use `#pragma push_macro` to save the macro before undefining it, and `#pragma pop_macro` to restore the macros and the end of the header. Reviewers: mclow.lists, bcraig, compnerd, EricWF Reviewed By: EricWF Subscribers: cfe-commits, krytarowski Differential Revision: https://reviews.llvm.org/D33080 llvm-svn: 304357
2017-06-01 06:07:49 +08:00
_LIBCPP_POP_MACROS
libcxx initial import llvm-svn: 103490
2010-05-12 03:42:16 +08:00
#endif // _LIBCPP_ISTREAM